Problems of the experiment, methods of its implementation. Typical errors and difficulties in the application of the experiment When conducting an experiment, it is impossible to avoid or reduce
![Problems of the experiment, methods of its implementation. Typical errors and difficulties in the application of the experiment When conducting an experiment, it is impossible to avoid or reduce](https://i2.wp.com/uhlib.ru/psihologija/yeksperimentalnaja_psihologija_konspekt_lekcii/_03.jpg)
V.V. Nikandrov points out that the achievement of the main goal of the experiment - the maximum possible unambiguity in understanding the connections between the phenomena of internal mental life and their external manifestations - is achieved due to the following main characteristics of the experiment:
1) the initiative of the experimenter in the manifestation of psychological facts of interest to him;
2) the possibility of varying the conditions for the emergence and development of mental phenomena;
3) strict control and fixation of the conditions and the process of their occurrence;
4) isolation of some and emphasis on other factors that determine the studied phenomena, which makes it possible to identify the patterns of their existence;
5) the possibility of repeating the conditions of the experiment for multiple verification of the obtained scientific data and their accumulation;
6) variation of conditions for quantitative assessments of the revealed regularities.
Thus, a psychological experiment can be defined as a method in which the researcher himself causes phenomena of interest to him and changes the conditions for their occurrence in order to establish the causes of these phenomena and the patterns of their development. In addition, the obtained scientific facts can be repeatedly reproduced due to the controllability and strict control of conditions, which makes it possible to verify them, as well as to accumulate quantitative data, on the basis of which it is possible to judge the typicality or randomness of the studied phenomena.
4.2. Types of psychological experiment
Experiments are of several types. Depending on the way of organizing distinguish laboratory, natural and field experiments. Laboratory The experiment is carried out under special conditions. The researcher deliberately and purposefully influences the object of study in order to change its state. The advantage of a laboratory experiment can be considered strict control over all conditions, as well as the use of special equipment for measurement. The disadvantage of a laboratory experiment is the difficulty of transferring the obtained data to real conditions. The subject in a laboratory experiment is always aware of his participation in it, which can cause motivational distortions.
Natural The experiment is carried out in real conditions. Its advantage lies in the fact that the study of the object is carried out in the context of everyday life, so the data obtained are easily transferred to reality. The subjects are not always informed about their participation in the experiment, so they do not give motivational distortions. Disadvantages - the inability to control all conditions, unforeseen interference and distortion.
Field The experiment is carried out according to the natural scheme. In this case, it is possible to use portable equipment, which makes it possible to more accurately record the received data. The subjects are informed about participation in the experiment, but the familiar environment reduces the level of motivational distortions.
Depending on the research objectives There are search, pilot and confirmatory experiments. Search the experiment is aimed at finding a cause-and-effect relationship between phenomena. It is carried out at the initial stage of the study, allows you to formulate a hypothesis, identify independent, dependent and side variables (see 4.4) and determine how to control them.
Aerobatic An experiment is a trial experiment, the first in a series. It is carried out on a small sample, without strict control of variables. The pilot experiment makes it possible to eliminate gross errors in the formulation of the hypothesis, to specify the goal, and to clarify the methodology for conducting the experiment.
Confirming the experiment is aimed at establishing the type of functional relationship and clarifying the quantitative relationships between variables. It is carried out at the final stage of the study.
Depending on the nature of influence on the subject allocate ascertaining, forming and control experiments. stating the experiment includes measuring the state of an object (a subject or a group of subjects) before active influence on it, diagnosing the initial state, establishing cause-and-effect relationships between phenomena. aim formative The experiment is the use of methods of active development or formation of any properties in the subjects. Control An experiment is a repeated measurement of the state of an object (subject or group of subjects) and comparison with the state before the start of the formative experiment, as well as with the state in which the control group is located, which did not receive experimental exposure.
By influence opportunities experimenter, the independent variable is allocated to the provoked experiment and the experiment to which they refer. provoked An experiment is an experiment in which the experimenter himself changes the independent variable, while the results observed by the experimenter (types of reactions of the subject) are considered provoked. P. Fress calls this type of experiment "classical". Experiment, which is referred to is an experiment in which changes in the independent variable are carried out without the intervention of the experimenter. This type of psychological experiment is resorted to when independent variables affect the subject, significantly extended in time (for example, the education system, etc.). If the impact on the subject can cause a serious negative physiological or psychological disturbance, then such an experiment cannot be carried out. However, there are cases when a negative impact (for example, a brain injury) occurs in reality. Subsequently, such cases can be generalized and studied.
4.3. Structure of a psychological experiment
The main components of any experiment are:
1) the subject (the subject or group under study);
2) experimenter (researcher);
3) stimulation (method of influence on the subject chosen by the experimenter);
4) the subject's response to stimulation (his mental reaction);
5) conditions of the experiment (additional to the stimulation of the impact, which can affect the reactions of the subject).
The response of the subject is an external reaction, by which one can judge the processes taking place in his inner, subjective space. These processes themselves are the result of the stimulation and conditions of experience acting on him.
If the response (reaction) of the subject is denoted by the symbol R, and the effects of the experimental situation on him (as a combination of stimulation effects and experimental conditions) - by the symbol S, then their ratio can be expressed by the formula R = =f (S). That is, the reaction is a function of the situation. But this formula does not take into account the active role of the psyche, the personality of a person. (P). In reality, a person's reaction to a situation is always mediated by the psyche, the personality. Thus, the relationship between the main elements of the experiment can be fixed by the following formula: R = f(R, S).
P. Fress and J. Piaget, depending on the objectives of the study, distinguish three classical types of relationships between these three components of the experiment: 1) functional relationships; 2) structural relations; 3) differential relations.
functional relationship characterized by the variability of responses (R) of the subject (P) with systematic qualitative or quantitative changes in the situation (S). Graphically, these relationships can be represented by the following diagram (Fig. 2).
Examples of Functional Relationships Identified in Experiments: Changing Feelings (R) depending on the intensity of the impact on the senses (S); storage capacity (R) on the number of repetitions (S); intensity of emotional response (R) on the action of various emotional factors (S); development of adaptation processes (R) in time (S) etc.
Structural relationships revealed through a system of responses (R1, R2, Rn) to various situations (Sv S2, Sn). Relationships between individual responses are structured into a system that reflects the personality structure (P). Schematically, it looks like this (Fig. 3).
![](https://i2.wp.com/uhlib.ru/psihologija/yeksperimentalnaja_psihologija_konspekt_lekcii/_03.jpg)
Examples of structural relationships: a system of emotional reactions (Rp R2, Rn) to the action of stressors (Sv S2, Sn); solution efficiency (R1, R2, Rn) various intellectual tasks (S1, S2, sn) etc.
Differential Relations revealed through reaction analysis (R1, R2, Rn) of different subjects (P1, P2, pn) for the same situation (S). The scheme of these relations is as follows (Fig. 4).
![](https://i2.wp.com/uhlib.ru/psihologija/yeksperimentalnaja_psihologija_konspekt_lekcii/_04.jpg)
Examples of differential relationships: the difference in the reaction speed of different people, national differences in the expressive manifestation of emotions, etc.
4.4. Experimental variables and how to control them
To clarify the ratio of all factors included in the experiment, the concept of "variable" is introduced. There are three types of variables: independent, dependent and additional.
Independent variables. The factor that is changed by the experimenter himself is called independent variable(NP).
The conditions under which the subject's activity is carried out, the characteristics of the tasks the performance of which is required from the subject, the characteristics of the subject himself (age, gender, and other differences in the subjects, emotional states and other properties of the subject or people interacting with him) can act as NP in the experiment. Therefore, it is customary to single out the following types NP: situational, instructive and personal.
situational NP most often are not included in the structure of the experimental task performed by the subject. Nevertheless, they have a direct impact on his activity and can be varied by the experimenter. Situational NPs include various physical parameters, such as illumination, temperature, noise level, as well as the size of the room, furnishings, placement of equipment, etc. The socio-psychological parameters of situational NPs can include performing an experimental task in isolation, in the presence of an experimenter, external observer or group of people. V.N. Druzhinin points to the features of communication and interaction between the subject and the experimenter as a special kind of situational NP. Much attention is paid to this aspect. In experimental psychology, there is a separate direction, which is called "psychology of psychological experiment".
Instructional NPs are directly related to the experimental task, its qualitative and quantitative characteristics, as well as the methods of its implementation. The instructive NP can be manipulated more or less freely by the experimenter. It can vary the material of the task (for example, numerical, verbal or figurative), the type of response of the subject (for example, verbal or non-verbal), the scale of assessment, etc. Great opportunities lie in the way in which the subjects are instructed, informing them about the purpose of the experimental task. The experimenter can change the means that are offered to the subject for completing the task, put obstacles in front of him, use a system of rewards and punishments in the course of completing the task, etc.
Personal NP are controlled features of the subject. Usually, such features are the states of the participant in the experiment, which the researcher can change, for example, various emotional states or states of performance-fatigue.
Each subject participating in the experiment has many unique physical, biological, psychological, socio-psychological and social characteristics that the experimenter cannot control. In some cases, these uncontrolled features should be considered additional variables and control methods should be applied to them, which will be discussed below. However, in differential psychological research, when using factorial designs, uncontrolled personal variables can act as one of the independent variables (for details on factorial designs, see 4.7).
Researchers also distinguish different kinds independent variables. Depending on the presentation scale qualitative and quantitative NPs can be distinguished. quality NPs correspond to different gradations of naming scales. For example, the subject's emotional states can be represented by states of joy, anger, fear, surprise, etc. Ways of performing tasks may include the presence or absence of prompts to the subject. quantitative NP correspond to rank, proportional or interval scales. For example, the time allotted to complete the task, the number of tasks, the amount of remuneration based on the results of solving problems can be used as quantitative NP.
Depending on the number of levels of manifestation independent variables distinguish two-level and multi-level NP. Two-level NPs have two levels of manifestation, multilevel- three or more levels. Depending on the number of levels of manifestation of NP, experimental plans of various complexity are built.
dependent variables. A factor whose change is a consequence of a change in the independent variable is called dependent variable(ZP). The dependent variable is the component of the subject's response that is of direct interest to the researcher. Physiological, emotional, behavioral reactions and other psychological characteristics that can be registered in the course of psychological experiments can act as RFP.
Depending on the the way in which changes can be registered, allocate ZP:
S observed directly;
S requiring physical equipment for measurement;
S requiring a psychological dimension.
To ZP, directly observable, include verbal and non-verbal behavioral manifestations that can be clearly and unambiguously assessed by an external observer, for example, refusal of an activity, crying, a certain statement of the subject, etc. physical equipment for registration, include physiological (pulse, blood pressure, etc.) and psychophysiological reactions (reaction time, latent time, duration, speed of actions, etc.). To RFP requiring psychological dimension, include such characteristics as the level of claims, the level of development or formation of certain qualities, forms of behavior, etc. For the psychological measurement of indicators, standardized procedures can be used - tests, questionnaires, etc. Some behavioral parameters can be measured, t i.e. unambiguously recognized and interpreted only by specially trained observers or experts.
Depending on the the number of parameters included in the dependent variable, one-dimensional, multidimensional and fundamental RFPs are distinguished. one-dimensional The RFP is represented by the only parameter whose changes are studied in the experiment. An example of a one-dimensional RFP is the speed of a sensorimotor reaction. Multidimensional ZP is represented by a set of parameters. For example, mindfulness can be measured by the amount of material viewed, the number of distractions, the number of correct and incorrect answers, etc. Each parameter can be recorded independently. Fundamental ZP is a variable of a complex nature, the parameters of which have certain known relationships with each other. In this case, some parameters act as arguments, and the dependent variable itself acts as a function. For example, the fundamental measurement of the level of aggression can be considered as a function of its individual manifestations (facial, verbal, physical, etc.).
The dependent variable must have such a basic characteristic as sensitivity. sensitivity ZP is its sensitivity to a change in the level of the independent variable. If the dependent variable does not change when the independent variable changes, then the latter is non-positive and it makes no sense to conduct an experiment in this case. There are two known variants of the manifestation of the insensitivity of the RFP: the “ceiling effect” and the “floor effect”. The "ceiling effect" is observed, for example, in the case when the task presented is so simple that it is performed by all subjects, regardless of age. The “gender effect”, on the other hand, occurs when the task is so difficult that none of the subjects can cope with it.
There are two main ways of fixing changes in BP in a psychological experiment: immediate and delayed. Direct the method is used, for example, in experiments on short-term memorization. The experimenter, immediately after repeating a series of stimuli, fixes their number reproduced by the subject. The delayed method is used when impact and the effect is a certain period of time (for example, when determining the influence of the number of memorized foreign words on the success of the translation of the text).
Additional variables(DP) is a concomitant stimulation of the subject that affects his response. The set of DP consists, as a rule, of two groups: external conditions of experience and internal factors. Accordingly, they are usually called external and internal DP. To external DP include the physical environment of the experiment (illumination, temperature, sound background, spatial characteristics of the room), parameters of apparatus and equipment (design of measuring instruments, operating noise, etc.), time parameters of the experiment (start time, duration, etc.), personality of the experimenter. To internal DP include the mood and motivation of the subjects, their attitude towards the experimenter and experiments, their psychological attitudes, inclinations, knowledge, skills, skills and experience in this type of activity, the level of fatigue, well-being, etc.
Ideally, the researcher seeks to reduce all additional variables to nothing, or at least to a minimum, in order to highlight the “pure” relationship between the independent and dependent variables. There are several main ways to control the influence of external DP: 1) elimination of external influences; 2) constancy of conditions; 3) balancing; 4) counterbalancing.
Elimination of external influences represents the most radical method of control. It consists in the complete exclusion from the external environment of any external DP. Conditions are created in the laboratory that isolate the test subject from sounds, light, vibration effects, etc. The most striking example is the sensory deprivation experiment conducted on volunteers in a special chamber that completely excludes any stimuli from the external environment. It should be noted that it is practically impossible to eliminate the effects of DP, and it is not always necessary, since the results obtained under the conditions of eliminating external influences can hardly be transferred to reality.
The next way to control is to create constant conditions. The essence of this method is to make the effects of DP constant and the same for all subjects throughout the experiment. In particular, the researcher strives to make constant the spatio-temporal conditions of the experiment, the technique of conducting it, the equipment, the presentation of instructions, etc. With careful application of this method of control, large errors can be avoided, however, the problem of transferring the results of the experiment to conditions that are very different from the experimental ones, remains problematic.
In cases where it is not possible to create and maintain constant conditions throughout the experiment, resort to the method balancing. This method is used, for example, in a situation where the external DP cannot be identified. In this case, balancing will consist in using the control group. The study of the control and experimental groups is carried out under the same conditions, with the only difference that in the control group there is no effect of the independent variable. Thus, the change in the dependent variable in the control group is due only to external DPs, while in the experimental group it is due to the combined action of external additional and independent variables.
If the external DP is known, then balancing consists in the effect of each of its values in combination with each level of the independent variable. In particular, such an external DP as the gender of the experimenter, in combination with the independent variable (gender of the subject), will lead to the creation of four experimental series:
1) male experimenter - male subjects;
2) male experimenter - female subjects;
3) female experimenter - male subjects;
4) female experimenter - female subjects.
In more complex experiments, balancing of several variables can be applied simultaneously.
counterbalancing as a way to control external DP is practiced most often when the experiment includes several series. The subject finds himself in different conditions sequentially, however, previous conditions may change the effect of subsequent ones. To eliminate the “sequence effect” that arises in this case, experimental conditions are presented to different groups of subjects in a different order. For example, in the first series of the experiment, the first group is presented with the solution of intellectual problems from simpler to more complex, and the second - from more complex to simpler. In the second series, on the contrary, the first group is presented with the solution of intellectual problems from more complex to simpler, and the second - from simpler to more complex. Counterbalancing is used in cases where it is possible to conduct several series of experiments, but it should be borne in mind that a large number of attempts causes fatigue for the subjects.
Internal DP, as mentioned above, are factors that lie in the personality of the subject. They have a very significant impact on the results of the experiment, their impact is quite difficult to control and take into account. Among the internal DP can be identified permanent and fickle. Permanent internal DPs do not change significantly during the experiment. If the experiment is conducted with one subject, then his gender, age, and nationality will be constant internal DP. This group of factors can also include the temperament, character, abilities, inclinations of the subject, his interests, views, beliefs and other components of the general orientation of the personality. In the case of an experiment with a group of subjects, these factors acquire the character of non-permanent internal DP, and then, to level their influence, they resort to special methods of forming experimental groups (see 4.6).
To fickle internal DP include the psychological and physiological characteristics of the subject, which can either change significantly during the experiment, or be updated (or disappear) depending on the goals, objectives, type, form of organization of the experiment. The first group of such factors consists of physiological and mental states, fatigue, addiction, the acquisition of experience and skills in the process of performing an experimental task. The other group includes the attitude towards this experience and this study, the level of motivation for this experimental activity, the attitude of the subject to the experimenter and his role as an experimental, etc.
To equalize the effect of these variables on responses in different samples, there are a number of methods that have been successfully used in experimental practice.
To eliminate the so-called serial effect, which is based on habituation, a special order of presentation of stimuli is used. This procedure is called "balanced alternating order", when stimuli of different categories are presented symmetrically with respect to the center of the stimulus row. The scheme of such a procedure looks like this: A B B A, where BUT and AT– incentives of different categories.
To prevent influence on the response of the subject anxiety or inexperience, conducting trial or preliminary experiments. Their totals are not taken into account when processing data.
To prevent variability in responses due to accumulation of experience and skills during the experiment, the subject is offered the so-called "exhaustive practice". As a result of this practice, the subject develops stable skills before the start of the actual experiment, and in further experiments, the subject's indicators do not directly depend on the factor of accumulating experience and skills.
In those cases where it is necessary to minimize the influence on the response of the subject fatigue, resort to the "rotation method". Its essence lies in the fact that each subgroup of subjects is presented with a certain combination of stimuli. The totality of such combinations completely exhausts the entire set of possible options. For example, with three types of stimuli (A, B, C), each of them is presented with the first, second and third place in the presentation to the subjects. Thus, stimuli are presented to the first subgroup in the order ABC, the second - AVB, the third - BAV, the fourth - BVA, the fifth - VAB, the sixth - VBA.
The above methods of procedural adjustment of internal non-constant DP are applicable both for individual and group experiments.
The set and motivation of the subjects as internal non-permanent DP must be maintained at the same level during the entire experiment. Installation how the readiness to perceive a stimulus and respond to it in a certain way is created through the instruction that the experimenter gives to the subject. In order for the installation to be exactly what is required for the task of the study, the instruction must be available to the subjects and adequate to the tasks of the experiment. The unambiguity and ease of understanding of the instruction are achieved by its clarity and simplicity. To avoid variability in presentation, it is recommended that instructions be read verbatim or given in writing. The maintenance of the initial set is controlled by the experimenter by constant observation of the subject and is corrected by recalling, if necessary, the appropriate instructions of the instruction.
Motivation The test subject is seen mainly as an interest in the experiment. If interest is absent or weak, then it is difficult to count on the completeness of the subjects' fulfillment of the tasks provided for in the experiment and on the reliability of his answers. Too high interest, "remotivation", is also fraught with inadequacy of the subject's answers. Therefore, in order to obtain an initially acceptable level of motivation, the experimenter must seriously approach the formation of the contingent of subjects and the selection of factors stimulating their motivation. Competitiveness, various types of remuneration, interest in one's performance, professional interest, etc. can serve as such factors.
Psychophysiological states it is recommended not only to keep the subjects at the same level, but also to optimize this level, i.e., the subjects must be in a “normal” state. You should make sure that before the experiment, the subject did not have super-significant experiences for him, he has enough time to participate in the experiment, he is not hungry, etc. During the experiment, the subject should not be unnecessarily excited or suppressed. If these conditions cannot be met, then it is better to postpone the experiment.
From the considered characteristics of variables and methods of their control, the need for careful preparation of the experiment during its planning becomes clear. In real conditions of experimentation, it is impossible to achieve 100% control of all variables, however, various psychological experiments differ significantly from each other in the degree of control of variables. The following section is devoted to the issue of assessing the quality of an experiment.
4.5. Validity and reliability of the experiment
For the design and evaluation of experimental procedures, the following concepts are used: an ideal experiment, an experiment of full compliance and an infinite experiment.
The Perfect Experiment is an experiment organized in such a way that the experimenter changes only the independent variable, the dependent variable is controlled, and all other conditions of the experiment remain unchanged. An ideal experiment assumes the equivalence of all subjects, the invariance of their characteristics over time, the absence of time itself. It can never be implemented in reality, since in life not only the parameters of interest to the researcher change, but also a number of other conditions.
The correspondence of a real experiment to an ideal one is expressed in such a characteristic as internal validity. Internal validity indicates the reliability of the results that a real experiment provides compared to an ideal one. The more dependent variables are affected by conditions not controlled by the researcher, the lower the internal validity of the experiment, therefore, the greater the likelihood that the facts found in the experiment are artifacts. High internal validity is the hallmark of a well-conducted experiment.
D. Campbell identifies the following factors that threaten the internal validity of the experiment: background factor, natural development factor, testing factor, measurement error, statistical regression, non-random selection, screening. If they are not controlled, then they lead to the appearance of the corresponding effects.
Factor background(stories) includes events that occur between the pre-measurement and the final measurement and may cause changes in the dependent variable along with the influence of the independent variable. Factor natural development due to the fact that changes in the level of the dependent variable may occur in connection with the natural development of the participants in the experiment (growing up, increasing fatigue, etc.). Factor testing lies in the influence of preliminary measurements on the results of subsequent ones. Factor measurement errors associated with inaccuracy or changes in the procedure or method of measuring the experimental effect. Factor statistical regression manifests itself in the event that subjects with extreme indicators of any assessments were selected for participation in the experiment. Factor non-random selection accordingly, it occurs in those cases when, when forming the sample, the selection of participants was carried out in a non-random manner. Factor sifting manifests itself in the event that the subjects drop out unevenly from the control and experimental groups.
The experimenter must take into account and, if possible, limit the influence of factors that threaten the internal validity of the experiment.
Full match experiment is an experimental study in which all conditions and their changes correspond to reality. The approximation of a real experiment to a full compliance experiment is expressed in terms of external validity. The degree of transferability of the results of the experiment to reality depends on the level of external validity. External validity, according to the definition of R. Gottsdanker, affects the reliability of the conclusions, which are given by the results of a real experiment compared to a full compliance experiment. To achieve high external validity, it is necessary that the levels of additional variables in the experiment correspond to their levels in reality. An experiment that lacks external validity is considered invalid.
Factors that threaten external validity include the following:
Reactive effect (consists in a decrease or increase in the susceptibility of subjects to experimental influence due to previous measurements);
The effect of the interaction of selection and influence (consists in the fact that the experimental influence will be significant only for the participants in this experiment);
Factor of experimental conditions (may lead to the fact that the experimental effect can be observed only in these specially organized conditions);
Interference factor of influences (appears when one group of subjects is presented with a sequence of mutually exclusive influences).
Care for the external validity of experiments is especially shown by researchers working in the applied fields of psychology - clinical, pedagogical, organizational, since in the case of an invalid study, its results will not give anything when transferred to real conditions.
Endless Experiment involves an unlimited number of experiments, samples to obtain more and more accurate results. An increase in the number of samples in an experiment with one subject leads to an increase reliability experiment results. In experiments with a group of subjects, an increase in reliability occurs with an increase in the number of subjects. However, the essence of the experiment lies precisely in the fact that, on the basis of a limited number of samples or with the help of a limited group of subjects, to identify causal relationships between phenomena. Therefore, an endless experiment is not only impossible, but also meaningless. To achieve high reliability of the experiment, the number of samples or the number of subjects must correspond to the variability of the phenomenon under study.
It should be noted that with an increase in the number of subjects, the external validity of the experiment also increases, since its results can be transferred to a wider population. To conduct experiments with a group of subjects, it is necessary to consider the issue of experimental samples.
4.6. Experimental samples
As mentioned above, the experiment can be carried out either with one subject or with a group of subjects. An experiment with one subject is carried out only in some specific situations. First, these are situations where the individual differences of the subjects can be neglected, i.e., any person can be the subject (if the experiment studies its features, unlike, for example, an animal). In other situations, on the contrary, the subject is a unique object (a brilliant chess player, musician, artist, etc.). There are also situations when the subject is required to have special competence as a result of training or extraordinary life experience (the only survivor in a plane crash, etc.). One test subject is also limited in cases where the repetition of this experiment with the participation of other subjects is impossible. For experiments with one subject, special experimental plans have been developed (for details, see 4.7).
More often experiments are carried out with a group of subjects. In these cases, the sample of subjects should be a model general population, to which the results of the study will then be extended. Initially, the researcher solves the problem of the size of the experimental sample. Depending on the purpose of the study and the possibility of the experimenter, it can range from several subjects to several thousand people. The number of subjects in a separate group (experimental or control) varies from 1 to 100 people. To apply statistical processing methods, it is recommended that the number of subjects in the compared groups be at least 30–35 people. In addition, it is advisable to increase the number of subjects by at least 5-10% of the required, since some of them or their results will be “rejected” during the experiment.
To form a sample of subjects, several criteria must be taken into account.
1. Informative. It lies in the fact that the selection of a group of subjects should correspond to the subject and hypothesis of the study. (For example, it makes no sense to recruit two-year-old children into a group of test subjects to determine the level of arbitrary memorization.) It is desirable to create ideal ideas about the object of experimental research and, when forming a group of test subjects, deviate minimally from the characteristics of the ideal experimental group.
2. Criterion of equivalence of subjects. When forming a group of subjects, one should take into account all the significant characteristics of the object of study, differences in the severity of which can significantly affect the dependent variable.
3. Representativeness criterion. The group of people participating in the experiment must represent the entire part of the general population to which the results of the experiment will apply. The size of the experimental sample is determined by the type of statistical measures and the chosen accuracy (reliability) of accepting or rejecting the experimental hypothesis.
Consider strategies for selecting subjects from a population.
Random Strategy is that each member of the general population is given an equal chance of being included in the experimental sample. To do this, each individual is assigned a number, and then an experimental sample is formed using a table of random numbers. This procedure is difficult to implement, since each representative of the population of interest to the researcher must be taken into account. In addition, the random strategy gives good results when forming a large experimental sample.
Stratometric selection is used in the event that the experimental sample must necessarily include subjects with a certain set of characteristics (gender, age, level of education, etc.). The sample is compiled in such a way that the subjects of each stratum (layer) with the given characteristics are equally represented in it.
Stratometric random selection combines the two previous strategies. Representatives of each stratum are assigned numbers and an experimental sample is randomly formed from them. This strategy is effective when selecting a small experimental sample.
Representative Modeling is used in the case when the researcher manages to create a model of an ideal object of experimental research. The characteristics of a real experimental sample should deviate minimally from the characteristics of an ideal experimental sample. If the researcher does not know all the characteristics of the ideal model of experimental research, then the strategy is applied approximate modeling. The more accurate the set of criteria that describe the population to which the conclusions of the experiment are supposed to be extended, the higher its external validity.
Sometimes, as an experimental sample, real groups, at the same time, either volunteers participate in the experiment, or all subjects are involved involuntarily. In both cases, external and internal validity are violated.
After the formation of the experimental sample, the experimenter draws up a research plan. Quite often, the experiment is carried out with several groups, experimental and control, which are placed in different conditions. The experimental and control groups should be equivalent at the start of the experimental exposure.
The procedure for selecting equivalent groups and subjects is called randomization. According to a number of authors, the equivalence of groups can be achieved by pairwise selection. In this case, the experimental and control groups are composed of individuals equivalent in terms of side parameters significant for the experiment. The ideal option for pairwise selection is to attract twin pairs. Randomization with stratification consists in the selection of homogeneous subgroups, in which the subjects are equalized in all characteristics, except for the additional variables of interest to the researcher. Sometimes, in order to highlight a significant additional variable, all subjects are tested and ranked according to the level of its severity. The experimental and control groups are formed so that subjects with the same or similar values of the variable fall into different groups. The distribution of subjects into experimental and control groups can be carried out and random method. As mentioned above, with a large number of experimental samples, this method gives quite satisfactory results.
4.7. Experimental plans
Experimental plan is a tactic of experimental research embodied in a specific system of experiment planning operations. The main criteria for classifying plans are:
Composition of participants (individual or group);
Number of independent variables and their levels;
Types of representation scales for independent variables;
Method of collecting experimental data;
Place and conditions of the experiment;
Features of the organization of the experimental impact and the method of control.
Plans for groups of subjects and for one subject. All experimental plans can be divided according to the composition of participants into plans for groups of subjects and plans for one subject.
Experiments with group of subjects have the following advantages: the possibility of generalizing the results of the experiment to the population; the possibility of using schemes of intergroup comparisons; saving time; application of methods of statistical analysis. The disadvantages of this type of experimental plans include: the impact of individual differences between people on the results of the experiment; the problem of the representativeness of the experimental sample; the problem of equivalence of groups of subjects.
Experiments with one test subject- this is a special case of "plans with a small N. J. Goodwin points to the following reasons for using such plans: the need for individual validity, since in experiments with large N a problem arises when the generalized data does not characterize any of the subjects. An experiment with one subject is also carried out in unique cases when, for a number of reasons, it is impossible to attract many participants. In these cases, the purpose of the experiment is to analyze unique phenomena and individual characteristics.
An experiment with a small N, according to D. Martin, has the following advantages: the absence of complex statistical calculations, the ease of interpreting the results, the possibility of studying unique cases, involving one or two participants, and ample opportunities for manipulating independent variables. It also has some disadvantages, in particular, the complexity of control procedures, the difficulty in generalizing the results; relative uneconomical time.
Consider plans for one subject.
Time series planning. The main indicator of the influence of the independent variable on the dependent one in the implementation of such a plan is the change in the nature of the responses of the subject over time. The simplest strategy: the scheme BUT– B. The subject initially performs activities under conditions A, and then under conditions B. To control the “placebo effect”, the following scheme is used: A - B - A.(“The placebo effect” is the reactions of the subjects to “empty” stimuli, corresponding to reactions to real stimuli.) In this case, the subject does not need to know in advance which of the conditions is “empty” and which is real. However, these schemes do not take into account the interaction of impacts, therefore, when planning time series, as a rule, regular alternation schemes are used (A - B - A– B), positional adjustment (А – B - B- A) or random alternation. The use of longer "long" time series increases the possibility of detecting the effect, but leads to a number of negative consequences - fatigue of the subject, reduced control over other additional variables, etc.
Alternative Impact Plan is a development of the time series plan. Its specificity lies in the fact that the impact BUT and AT randomly distributed over time and presented to the subject separately. Then the effects of each of the exposures are compared.
Reverse plan used to study two alternative forms of behavior. Initially, the basic level of manifestation of both forms of behavior is recorded. Then a complex effect is presented, consisting of a specific component for the first form of behavior and an additional one for the second. After a certain time, the combination of influences is modified. The effect of two complex impacts is evaluated.
Criteria Increasing Plan often used in the psychology of learning. Its essence lies in the fact that a change in the behavior of the subject is recorded in response to an increase in exposure. In this case, the next impact is presented only after the subject reaches the given level of the criterion.
When conducting experiments with one subject, it should be taken into account that the main artifacts are practically irremovable. In addition, in this case, as in no other, the influence of the experimenter's attitudes and the relationship that develops between him and the subject is manifested.
R. Gottsdanker proposes to distinguish qualitative and quantitative experimental designs. AT quality In plans, the independent variable is presented on a nominative scale, i.e., two or more qualitatively different conditions are used in the experiment.
AT quantitative experimental plans, the levels of the independent variable are presented in interval, rank, or proportional scales, i.e., the levels of severity of a particular condition are used in the experiment.
A situation is possible when in a factorial experiment one variable will be presented in a quantitative form, and the other in a qualitative form. In this case, the plan will be combined.
Intragroup and intergroup experimental plans. T.V. Kornilova defines two types of experimental plans according to the criterion of the number of groups and the conditions of the experiment: intragroup and intergroup. To intragroup include designs in which the influence of variants of the independent variable and the measurement of the experimental effect occur in the same group. AT intergroup plans, the influence of variants of the independent variable is carried out in different experimental groups.
The advantages of the intragroup plan are: a smaller number of participants, the elimination of factors of individual differences, a decrease in the total time of the experiment, the possibility of proving the statistical significance of the experimental effect. Disadvantages include non-constancy of conditions and manifestation of the “sequence effect”.
The advantages of the intergroup design are: the absence of a "consistency effect", the possibility of obtaining more data, reducing the time of participation in the experiment for each subject, reducing the effect of dropping out of the experiment participants. The main disadvantage of the intergroup plan is the non-equivalence of groups.
Designs with one independent variable and factorial designs. According to the criterion of the number of experimental influences, D. Martin proposes to distinguish between plans with one independent variable, factorial plans and plans with a series of experiments. In the plans with one independent variable the experimenter manipulates one independent variable, which can have an unlimited number of manifestations. AT factorial plans (for details on them, see p. 120), the experimenter manipulates two or more independent variables, explores all possible options for the interaction of their different levels.
Plans from a series of experiments conducted to gradually exclude competing hypotheses. At the end of the series, the experimenter comes to the verification of one hypothesis.
Pre-experimental, quasi-experimental and true experimental designs. D. Campbell proposed to divide all experimental plans for groups of subjects into the following groups: pre-experimental, quasi-experimental and plans for true experiments. This division is based on the closeness of a real experiment to an ideal one. The fewer artifacts a particular plan provokes and the stricter the control of additional variables, the closer the experiment is to the ideal. Pre-experimental plans least of all take into account the requirements for an ideal experiment. V.N. Druzhinin points out that they can only serve as an illustration, in the practice of scientific research they should be avoided if possible. Quasi-experimental plans are an attempt to take into account the realities of life when conducting empirical research, they are specially created with a deviation from the schemes of true experiments. The researcher must be aware of the sources of artifacts - external additional variables that he cannot control. A quasi-experimental plan is used when a better plan cannot be applied.
Systematized signs of pre-experimental, quasi-experimental plans and plans of true experiments are given in the table below.
![](https://i0.wp.com/uhlib.ru/psihologija/yeksperimentalnaja_psihologija_konspekt_lekcii/_05.jpg)
When describing the experimental plans, we will use the symbolization proposed by D. Campbell: R- randomization; X– experimental impact; O- testing.
To pre-experimental plans include: 1) study of a single case; 2) a plan with preliminary and final testing of one group; 3) comparison of statistical groups.
At case study one group is tested once after the experimental exposure. Schematically, this plan can be written as:
The control of external variables and the independent variable is completely absent. In such an experiment, there is no material for comparison. The results can only be compared with ordinary ideas about reality; they do not carry scientific information.
Plan with preliminary and final testing of one group often used in sociological, socio-psychological and pedagogical research. It can be written as:
There is no control group in this plan, so it cannot be argued that changes in the dependent variable (difference between O1 and O2) recorded during testing are caused precisely by the change in the independent variable. Between the initial and final testing, other "background" events may occur that affect the subjects along with the independent variable. This plan also does not allow control over the effect of natural development and the effect of testing.
Comparison of statistical groups it would be more accurate to call it a plan for two non-equivalent groups with post-exposure testing. It can be written like this:
This plan takes into account the effect of testing by introducing a control group to control a number of external variables. However, with its help it is impossible to take into account the effect of natural development, since there is no material for comparing the state of the subjects at the moment with their initial state (no preliminary testing was carried out). To compare the results of the control and experimental groups, Student's t-test is used. However, it should be borne in mind that differences in test results may not be due to experimental exposure, but to differences in the composition of the groups.
Quasi-experimental plans are a kind of compromise between reality and the strict framework of true experiments. There are the following types of quasi-experimental plans in psychological research: 1) plans for experiments for non-equivalent groups; 2) plans with preliminary and final testing of various randomized groups; 3) plans for discrete time series.
Plan experiment for nonequivalent groups is aimed at establishing a causal relationship between variables, however, it lacks a procedure for equalizing groups (randomization). This plan can be represented by the following diagram:
In this case, two real groups are involved in the experiment. Both groups are being tested. Then one group is subjected to experimental treatment and the other is not. Both groups are then retested. The results of the first and second testing of both groups are compared, for comparison, Student's t-test and analysis of variance are used. Difference O2 and O4 indicates natural development and background exposure. To identify the effect of an independent variable, it is necessary to compare 6(O1 O2) and 6(O3 O4), i.e., the magnitude of the shifts in indicators. The significance of the difference in the growth of indicators will indicate the influence of the independent variable on the dependent one. This design is similar to the true two-group experiment with pre- and post-exposure testing (see p. 118). The main source of artifacts is the difference in the composition of groups.
Plan with pre and post testing of various randomized groups differs from the design of a true experiment in that one group passes the preliminary test, and the final test is the equivalent group that was exposed to:
The main disadvantage of this quasi-experimental design is the inability to control the "background" effect - the influence of events that occur along with the experimental exposure in the period between the first and second testing.
Plans discrete time series are subdivided into several types depending on the number of groups (one or more), as well as depending on the number of experimental effects (single or series of effects).
The plan of discrete time series for one group of subjects is that the initial level of the dependent variable on the group of subjects is initially determined using a series of consecutive measurements. Then an experimental effect is applied and a series of similar measurements is carried out. Compare the levels of the dependent variable before and after exposure. Schematic of this plan:
The main disadvantage of the discrete time series design is that it does not allow one to separate the effect of the influence of the independent variable from the influence of the background events that occur during the study.
A modification of this design is a time-series quasi-experiment in which pre-measurement exposure alternates with no pre-measurement exposure. His schema is:
XO1 - O2XO3 - O4 XO5
Alternation can be regular or random. This option is only suitable if the effect is reversible. When processing the data obtained in the experiment, the series are divided into two sequences and the results of measurements, where there was an impact, are compared with the results of measurements, where it was absent. To compare data, Student's t-test is used with the number of degrees of freedom n– 2, where n is the number of situations of the same type.
Time series plans are often implemented in practice. However, when using them, the so-called "Hawthorne effect" is often observed. It was first discovered by American scientists in 1939, when they were conducting research at the Hawthorne plant in Chicago. It was assumed that the change in the system of labor organization would increase its productivity. However, during the experiment, any changes in the organization of labor led to an increase in its productivity. As a result, it turned out that participation in the experiment itself increased the motivation to work. The subjects realized that they were personally interested in them, and began to work more productively. To control for this effect, a control group must be used.
The scheme of the time series plan for two non-equivalent groups, of which one is not affected, looks like this:
O1O2O3O4O5O6O7O8O9O10
O1O2O3O4O5O6O7O8O9O10
Such a plan allows you to control the "background" effect. It is usually used by researchers when studying real groups in educational institutions, clinics, and in production.
Another specific plan, which is often used in psychology, is called an experiment. ex post facto. It is often used in sociology, pedagogy, as well as in neuropsychology and clinical psychology. The strategy for implementing this plan is as follows. The experimenter himself does not influence the subjects. Some real event from their life acts as an influence. The experimental group consists of "subjects" who have been exposed, while the control group consists of people who have not experienced it. In this case, the groups, if possible, are equalized at the moment of their state before the impact. Then the dependent variable is tested in the representatives of the experimental and control groups. The data obtained as a result of testing are compared and a conclusion is made about the impact of exposure on the further behavior of the subjects. Thus the plan ex post facto simulates the design of the experiment for two groups with their equalization and testing after exposure. His schema is:
If it is possible to achieve group equivalence, then this design becomes the design of a true experiment. It is implemented in many modern studies. For example, in the study of post-traumatic stress, when people who have suffered the effects of a natural or man-made disaster, or combatants are tested for the presence of post-traumatic stress syndrome, their results are compared with the results of the control group, which makes it possible to identify the mechanisms for the occurrence of such reactions. In the neuropsychology of brain injury, lesions of certain structures, considered as "experimental exposure", provide a unique opportunity to identify the localization of mental functions.
Plans for true experiments for one independent variable differ from others as follows:
1) using strategies for creating equivalent groups (randomization);
2) the presence of at least one experimental and one control group;
3) final testing and comparison of the results of groups that received and did not receive exposure.
Let us consider in more detail some experimental designs for one independent variable.
Plan for two randomized groups with post-exposure testing. His schema looks like this:
This plan is used if it is not possible or necessary to conduct preliminary testing. When the experimental and control groups are equal, this plan is the best, since it allows you to control most of the sources of artifacts. The absence of preliminary testing excludes both the effect of the interaction of the testing procedure and the experimental task, and the effect of testing itself. The plan allows you to control the influence of the composition of groups, spontaneous dropout, the influence of the background and natural development, the interaction of the composition of the group with other factors.
In the considered example, one level of influence of the independent variable was used. If it has several levels, then the number of experimental groups increases to the number of levels of the independent variable.
Plan for two randomized groups with pre and post testing. The outline of the plan looks like this:
R O1 X O2
This plan is used when there is doubt about the results of randomization. The main source of artifacts is the interaction between testing and experimental exposure. In reality, one also has to deal with the effect of testing non-simultaneity. Therefore, it is considered best to conduct testing of members of the experimental and control groups in random order. Presentation-non-presentation of the experimental impact is also best done in a random order. D. Campbell notes the need to control "intragroup events". This experimental design controls well the background effect and the natural development effect.
When processing data, parametric criteria are usually used. t and F(for data on an interval scale). Three values of t are calculated: 1) between O1 and O2; 2) between O3 and O4; 3) between O2 and O4. The hypothesis of the significance of the influence of the independent variable on the dependent variable can be accepted if two conditions are met: 1) differences between O1 and O2 important, and between O3 and O4 insignificant and 2) differences between O2 and O4 significant. Sometimes it is more convenient to compare not the absolute values, but the increments of the indicators b(1 2) and b(3 4). These values are also compared by Student's t-test. If the differences are significant, an experimental hypothesis is accepted about the influence of the independent variable on the dependent one.
Solomon's plan is a combination of the two previous plans. For its implementation, two experimental (E) and two control (C) groups are required. His schema looks like this:
![](https://i0.wp.com/uhlib.ru/psihologija/yeksperimentalnaja_psihologija_konspekt_lekcii/_06.jpg)
With this plan, the interaction effect of pre-testing and the experimental exposure effect can be controlled. The effect of experimental exposure is revealed by comparing the indicators: O1 and O2; O2 and O4; O5 and O6; O5 and O3. Comparison of O6, O1 and O3 reveals the effect of natural development and background influences on the dependent variable.
Now consider a design for one independent variable and several groups.
Design for three randomized groups and three levels of the independent variable used in cases where it is necessary to identify quantitative relationships between the independent and dependent variables. His schema looks like this:
![](https://i0.wp.com/uhlib.ru/psihologija/yeksperimentalnaja_psihologija_konspekt_lekcii/_07.jpg)
When implementing this plan, each group is presented with only one level of the independent variable. If necessary, you can increase the number of experimental groups in accordance with the number of levels of the independent variable. All of the above statistical methods can be used to process the data obtained with such an experimental design.
Factorial Experimental Designs are used to test complex hypotheses about relationships between variables. In a factorial experiment, as a rule, two types of hypotheses are tested: 1) hypotheses about the separate influence of each of the independent variables; 2) hypotheses about the interaction of variables. The factorial design is to ensure that all levels of independent variables are combined with each other. The number of experimental groups is equal to the number of combinations.
Factorial design for two independent variables and two levels (2 x 2). This is the simplest of factorial designs. His diagram looks like this.
![](https://i1.wp.com/uhlib.ru/psihologija/yeksperimentalnaja_psihologija_konspekt_lekcii/_08.jpg)
This plan reveals the effect of two independent variables on one dependent variable. The experimenter combines possible variables and levels. Sometimes four independent randomized experimental groups are used. Fisher's analysis of variance is used to process the results.
There are more complex versions of the factorial design: 3 x 2 and 3 x 3, etc. The addition of each level of the independent variable increases the number of experimental groups.
"Latin Square". It is a simplification of the full plan for three independent variables with two or more levels. The principle of the Latin square is that two levels of different variables occur only once in the experimental plan. This significantly reduces the number of groups and the experimental sample as a whole.
For example, for three independent variables (L, M, N) with three levels each (1, 2, 3 and N(A, B, C)) the plan according to the "Latin square" method will look like this.
![](https://i0.wp.com/uhlib.ru/psihologija/yeksperimentalnaja_psihologija_konspekt_lekcii/_09.jpg)
In this case, the level of the third independent variable (A, B, C) occurs in each line and in each column once. By combining the results across rows, columns, and levels, it is possible to identify the influence of each of the independent variables on the dependent variable, as well as the degree of pairwise interaction of the variables. The use of Latin letters A, B, FROM It is traditional to designate the levels of the third variable, which is why the method was called the “Latin square”.
"Greco-Latin square". This plan is used when it is necessary to investigate the influence of four independent variables. It is built on the basis of a Latin square for three variables, with a Greek letter attached to each Latin group of the plan, denoting the levels of the fourth variable. The schema for a plan with four independent variables, each with three levels, would look like this:
![](https://i2.wp.com/uhlib.ru/psihologija/yeksperimentalnaja_psihologija_konspekt_lekcii/_10.jpg)
To process the data obtained in terms of the "Greek-Latin square", the method of variance analysis according to Fisher is used.
The main problem that factorial designs can solve is determining the interaction of two or more variables. This problem cannot be solved by applying several conventional experiments with one independent variable. In the factorial plan, instead of trying to “clear” the experimental situation of additional variables (with a threat to external validity), the experimenter brings it closer to reality by introducing some additional variables into the category of independent ones. At the same time, the analysis of the relationships between the studied characteristics allows us to reveal hidden structural factors on which the parameters of the measured variable depend.
4.8. Correlation Studies
The theory of correlation research was developed by the English mathematician K. Pearson. The strategy for conducting such a study is that there is no controlled impact on the object. The plan of the correlation study is simple. The researcher puts forward a hypothesis about the presence of a statistical relationship between several mental properties of an individual. However, the assumption of causal dependence is not discussed.
Correlative is a study conducted to confirm or refute the hypothesis of a statistical relationship between several (two or more) variables. In psychology, mental properties, processes, states, etc. can act as variables.
Correlations.“Correlation” literally means ratio. If a change in one variable is accompanied by a change in another, then we speak of the correlation of these variables. The presence of a correlation between two variables is not evidence of the presence of causal relationships between them, but it makes it possible to put forward such a hypothesis. The absence of correlation allows one to refute the hypothesis of a causal relationship of variables.
There are several types of correlations:
Direct correlation (the level of one variable directly corresponds to the level of another variable);
Correlation due to a third variable (the level of one variable corresponds to the level of another variable due to the fact that both of these variables are due to a third, common variable);
Random correlation (not due to any variable);
Correlation due to the heterogeneity of the sample (if the sample consists of two heterogeneous groups, then a correlation can be obtained that does not exist in the general population).
Correlations are of the following types:
– positive correlation (an increase in the level of one variable is accompanied by an increase in the level of another variable);
– negative correlation (an increase in the level of one variable is accompanied by a decrease in the level of another);
- zero correlation (indicates the absence of a connection between variables);
- non-linear relationship (within certain limits, an increase in the level of one variable is accompanied by an increase in the level of another, and with other parameters - vice versa. Most psychological variables have a non-linear relationship).
Planning a correlation study. The design of the correlation study is a kind of quasi-experimental design in the absence of the influence of the independent variable on the dependent ones. A correlation study is broken down into a series of independent measurements in a group of subjects. When simple correlation study group is homogeneous. When comparative correlation study, we have several subgroups that differ in one or more criteria. The results of such measurements give a matrix of the form R x O. Correlation study data is processed by calculating correlations by rows or columns of the matrix. Row correlation yields a comparison of subjects. Column correlation provides information about the association of measured variables. Temporal correlations are often detected, i.e., changes in the structure of correlations over time.
The main types of correlation research are considered below.
Comparison of two groups. It is used to establish the similarity or difference between two natural or randomized groups in terms of the severity of one or another parameter. The mean results of the two groups are compared using Student's t-test. If necessary, Fisher's t-test (see 7.3) can also be used to compare the variances of an indicator between two groups.
Univariate study of one group under different conditions. The design of this study is close to experimental. But in the case of a correlation study, we do not control the independent variable, but only state the change in the individual's behavior under different conditions.
Correlation study of pairwise equivalent groups. This plan is used in the study of twins by the method of intra-pair correlations. The twin method is based on the following provisions: the genotypes of monozygotic twins are 100% similar, and dizygotic twins are 50% similar, the development environment for both dizygotic and monozygotic pairs is the same. Dizygotic and monozygotic twins are divided into groups: each contains one twin from a pair. In twins of both groups, the parameter of interest to the researcher is measured. Then the correlations between the parameters are calculated (O-correlation) and between twins (R-correlation). Comparing the intra-pair correlations of monozygotic and dizygotic twins, it is possible to identify the shares of the influence of the environment and the genotype on the development of a particular trait. If the correlation of monozygotic twins is reliably higher than the correlation of dizygotic twins, then we can talk about the existing genetic determination of the trait, otherwise we talk about environmental determination.
Multivariate correlation study. It is carried out to test the hypothesis about the relationship of several variables. An experimental group is selected, which is tested according to a specific program consisting of several tests. Research data are entered in the table of "raw" data. Then this table is processed, the coefficients of linear correlations are calculated. Correlations are evaluated for statistical differences.
Structural correlation study. The researcher reveals the difference in the level of correlation dependencies between the same indicators measured in representatives of different groups.
Longitudinal correlation study. It is built according to the plan of time series with testing of the group at specified intervals. In contrast to a simple longitudinal, the researcher is interested in changes not so much in the variables themselves as in the relationships between them.
General principles for designing experiments
Comparison.
Randomization.
Replication.
Uniformity.
Stratification.
factor levels
Title: General Principles of Designing Experiments
Detailed description:
Since its inception, science has been looking for ways to understand the laws of the surrounding world. Making one discovery after another, scientists rise higher and higher up the ladder of knowledge, erasing the border of the unknown and entering new frontiers of science. This way lies through experiment. Consciously limiting the infinite diversity of nature by the artificial framework of scientific experience, we turn it into a picture of the world that is understandable to the human mind.
Experiment as scientific research is the form in which and through which science exists and develops. The experiment requires careful preparation before it is carried out. In biomedical research, the planning of the experimental part of the study is of particular importance due to the wide variability of properties characteristic of biological objects. This feature is the main reason for the difficulties in interpreting the results, which can vary significantly from experience to experience.
Statistical problems justify the need to choose such an experimental scheme that would minimize the effect of variability on the scientist's conclusions. Therefore, the purpose of experiment design is to create a design that is necessary to obtain as much information as possible at the lowest cost to carry out the study. More precisely, the planning of an experiment can be defined as the procedure for choosing the number and conditions for conducting experiments that are necessary and sufficient to solve the problem with the required accuracy.
Experimental design originated in agrobiology and is associated with the English statistician and biologist Sir Ronald Aylmer Fisher. At the beginning of the 20th century, at the agrobiological station in Rothamsted (Great Britain), studies began on the effect of fertilizers on the yield of various cereal varieties. Scientists had to take into account both the great variability of the objects of study and the long duration of the experiments (about a year). Under these conditions, there was no other way but to develop a well-thought-out experimental plan to reduce the negative impact of these factors on the accuracy of the conclusions. Applying statistical knowledge to biological problems, Fisher came to develop his own principles of the theory of statistical inference and laid the foundation for a new science of planning and analyzing experiments.
Ronald Fisher himself explained the basics of planning on the example of an experiment carried out to determine the ability of a certain English lady to distinguish between what was poured into a cup in the first place - tea or milk. It should be noted that for real English ladies it is important that tea is poured into milk, and not vice versa, a violation of the sequence will be a sign of ignorance and spoil the taste of the drink.
The experiment is simple: the lady tastes tea with milk and tries to understand the order in which both ingredients were poured. The design developed for this study has a number of properties.
Comparison. In many studies, it is difficult or impossible to accurately determine the measurement result. So, for example, a lady will not be able to quantify the quality of tea, she will compare it with the standard of a properly prepared drink, the taste of which has been familiar to her since childhood. As a rule, in a scientific experiment, the object is compared either with some predetermined standard or with a control object.
Randomization. This is a very important point in planning. In our example, randomization refers to the order in which the cups are presented for tasting. Randomization is necessary in order to be able to use statistical methods to analyze the results of the study.
Replication. Repeatability is a necessary component of setting up an experiment. It is unacceptable to draw conclusions about the ability to determine the quality of tea from only one cup. The result of each individual measurement (tasting) carries a share of the uncertainty that has arisen under the influence of many random factors. Therefore, several tests are needed to identify the source of variability. The sensitivity of the experiment is related to this property. Fisher noted that until the number of cups of tea exceeds a certain minimum, it is impossible to draw any unambiguous conclusions.
Uniformity. Despite the need to repeat measurements (replication), their number should not be too large so as not to lose homogeneity. The temperature difference of the cups, dullness of taste, etc., when a certain limit number of repetitions is exceeded, can make it difficult to analyze the results of the experiment.
Stratification. Going beyond R. Fisher's example to a more abstract description of the experimental plan, one can additionally indicate such a property as stratification (blocking). Stratification is the distribution of experimental units into relatively homogeneous groups (blocks, layers). The stratification procedure allows minimizing the effect of non-random sources of variability known to us. Within each block, the experimental error is assumed to be smaller relative to the variant with random selection for the experiment of the same number of objects. For example, in a study of a new drug, we have two levels of the factor, "drug" and "placebo", which are given to men and women. In this case, gender is a blocking factor, according to which the study is divided into subgroups.
The characteristics of an experimental design described above apply in whole or in part to any scientific experiment. However, to get started, it is not enough just to know about the general properties of the study; more thorough preparation is needed. Creating a detailed guide within the framework of one article is impossible, therefore, the most general information about the stages of planning an experiment will be presented here.
Any research begins with setting a goal. The choice of problem to study and its formulation will influence both the design of the study and the conclusions that will be drawn from its results. In the simplest case, the problem statement should include the questions “Who?”, “What?”, “When?”, “Why?” And How?".
An illustration of the importance of this planning stage can be found in a study that collects information on traffic accidents. Depending on the goal setting, the work can be directed to the development of a new car or a new road surface. Despite the fact that the same data set is used, the problem statement and conclusions differ significantly depending on the problem formulation.
After choosing the goal of the work, the so-called dependent variables should be determined. These are the variables that will be measured in the study. For example, indicators of the functioning of certain systems of the human body or laboratory animals (heart rate, blood pressure, enzyme levels in the blood, etc.), as well as any other characteristics of the objects of study, the change of which will be informative for us.
Since there are dependent variables, there must also be independent variables. Their other name is factors. The researcher operates with factors in the experiment. This may be the dose of the study drug, stress level, degree of exercise, etc. The relationship between a factor and a dependent variable is conveniently represented using a cybernetic system, often referred to as a "black box".
A black box is a system whose mechanism of operation is unknown to us. However, the researcher has information about what happens at the input and output of the black box. The state of the output is functionally dependent on the state of the input. Accordingly, y1, y2, ..., yp are dependent variables, the value of which depends on factors (independent variables x1, x2, ..., xk). Parameters w1, w2, ..., wn are disturbing influences that cannot be controlled or change with time.
In general terms, this can be written as follows: y=f(x1, x2, ..., xk).
Each factor in the experience can take on one of several values. Such values are called factor levels. It may turn out that the factor is capable of taking an infinite number of values (for example, the dose of a drug), but in practice several discrete levels are chosen, the number of which depends on the objectives of a particular experiment.
A fixed set of factor levels defines one of the possible states of the black box. At the same time, these are the conditions for conducting one of the possible experiments. If we enumerate all possible sets of such states, then we will get a complete set of different states of the given system, the number of which will be the number of all possible experiments. In order to calculate the number of possible states, it is enough to raise the number of levels of factors q (if it is the same for all factors) to the power of the number of factors k.
The totality of all possible states determines the complexity of the black box. Thus, a system of ten factors at four levels can be in more than a million different states. Obviously, in such cases it is impossible to carry out a study that includes all possible experiments. Therefore, at the planning stage, the question of how many experiments and which ones need to be carried out to solve the problem is decided.
It should be noted that the properties of the object of study are essential for the experiment. First, we need to have information about the degree of reproducibility of the results of experiments with a given object. To do this, you can conduct an experiment, and then repeat it at irregular intervals and compare the results. If the spread of values does not exceed our requirements for the accuracy of the experiment, then the object satisfies the requirement for reproducibility of results. Another requirement for an object is its manageability. A controllable object is an object on which an active experiment can be carried out. In turn, an active experiment is an experiment in which the researcher has the opportunity to choose the levels of factors that are of interest to him.
In practice, there are no fully managed objects. As mentioned above, both controllable and uncontrollable factors act on a real object, which leads to variability in the results between individual objects. We can separate random changes from regular ones, caused by different levels of independent variables, only with the help of statistical methods.
But statistical methods are effective only under certain conditions. One of these conditions is the requirement of a certain minimum sample size used in the experiment. It is obvious that the wider the range of change in attributes from object to object, the greater should be the repetition of the experiment, i.e., the number of experimental groups.
Since an unreasonably large number of trials will make the study too expensive, and an insufficient sample size may compromise the accuracy of the conclusions, determining the required sample size plays a crucial role in experiment design. Methods for calculating the minimum sample size are described in detail in the specialized literature, so it is not possible to present them in the article. However, it should be mentioned that they require a preliminary determination of the average value of the indicator under study and its error. Publications about similar studies can serve as a source of such information. If they have not yet been carried out, then there is a need to perform a preliminary “pilot” study to assess the variability of the trait.
The next step in designing experiments is randomization. Randomization is a process used to group subjects so that each of them has an equal chance of being placed in a control or treatment group. In other words, the selection of study participants must be random so that the study is not biased towards the investigator's "preferred" outcome.
Randomization helps prevent bias due to causes that were not directly addressed in the experimental design. For this, for example, the formation of experimental groups of laboratory animals is carried out randomly. However, complete randomization is not always possible. Thus, patients of a certain age group, with a predetermined diagnosis and severity of the disease, take part in clinical trials, and, therefore, the selection of participants is not random. In addition, randomization is limited by the so-called "block" designs of experiments. These plans imply that selection in each block is carried out in accordance with certain non-random conditions, and random selection of research objects is possible only within blocks. The randomization process is easy to implement using specialized statistical software or special tables.
In conclusion, it is necessary to say about the need to take into account in the research plan, in addition to the requirements of medicine and statistics, also moral and ethical standards. Do not forget that not only people, but also laboratory animals should be involved in the experiment in accordance with ethical principles.
A scientific experiment is a research method that provides a scientifically objective verification of the correctness of the hypothesis justified at the beginning of the study. The experiment makes it possible to detect recurring, stable, necessary, essential connections between phenomena, i.e. study the patterns that characterize any process or phenomena. Unlike observation, the experiment allows one to artificially separate the phenomenon under study from others, purposefully change the conditions for its implementation. At the same time, the experiment requires the researcher to have a higher level of training, mastery of the methodology for setting up and conducting an experiment, and the ability to develop an experiment program.
In research activities, different types of experiments are used. The most common laboratory and natural experiment. In the first case, the experiment is carried out in specially prepared conditions - a laboratory, where an object is isolated from a complex system of relationships, which are replaced by specially simulated conditions. For example, natural heating replaces artificial heating, and other conditions are also modeled: illumination, pressure, mechanical effects, etc.
A natural experiment is carried out in ordinary, natural conditions, where the experimenter observes the initial state of the object, its development and disappearance. In this case, the object can be subjected to a certain influence on the part of the experimenter. Then the whole process is repeated, for example, the resettlement and acclimatization of plants or animals.
When conducting an experiment, it is necessary to carry out a representative (indicative for the entire population) sample of the number of experimental objects.
The sample should be representative in terms of coverage of the participants in the experiment. For example, when conducting an experiment in the social sphere, it is necessary to represent all groups of the population,
if the goal of this experiment will receive a result that affects the entire society. Sometimes the theme of the experiment allows us to confine ourselves to a laboratory study, for example, a high-quality express method for detecting heavy metal cations in drinking water.
Thus, there is not and cannot be any template decision on the choice of the number of experimental objects, but the representativeness of the sample must always be proven from the point of view of the objectivity of the results obtained. When conducting an educational study, it is impossible to achieve the optimal ratio of the number of objects selected for the experiment. As a rule, it is always underestimated, but given that the didactic task of teaching students is on a different plane than a purely research task, one can also rely on a smaller sample. The same applies to determining the required duration of the experiment. Its too short period leads to biased scientific data, too long - increases the complexity and is unacceptable from the point of view of completeness (for the student, this is the time of studying at school).
Therefore, it is desirable for each researcher to justify the duration of the experiment. This can be done, firstly, by analyzing the previous experience of similar experiments in which correct scientific and practical conclusions were made; secondly, by correlating the goals and objectives of the experiment with its required duration.
Example. 1. When studying the features of bird nesting, the experiment will last the entire period during which birds build nests and lay eggs.
If during the experiment the influence of any substances (conditions) on the manifestation of certain regularities is studied, then it is necessary to cover the most typical regularities in the experiment.
2. When conducting an experiment to determine the "impact of noise on the performance of students", its duration
cannot be limited to 1-2 days or one source of noise (industrial, non-industrial). The duration of the specified experiment should be within at least an academic year. If the effect of fertilization on the yield of variety X or maturity is being studied, such an experiment usually lasts more than one year.
Conducting an experiment requires the choice of a specific technique. This is preceded by work on the study of the initial level of the state of the experimental object. So, analyzing the experiment on studying the state of the moss-lichen cover of the biocenosis, it is necessary to make sure that in this biocenosis, mosses and lichens are not represented by one or two species, but occupy a whole ecological niche.
For each specific case, not the entire set of known methods is chosen, but such a combination of them that will give reliable information. For example, when determining the MPC of copper in water, it is necessary to use the technique of both qualitative and quantitative detection.
Experimental activity implies the presence of a control object, which is a criterion for evaluating the results of the experiment. For example, when conducting an experiment on the effect of fertilizers on the ripening time, there must be a control plot on which fertilizer was not applied. When determining the MPC content of copper in water, it is necessary to have reliable figures for MPC (1.1 mg/l).
The experiment requires keeping a record, in which the facts of experimental activity are entered using text, numbers, symbols, schemes. As already noted, the protocol must be consistent, consistent and adequate, i.e., allowing conclusions to be drawn on the basis of objective information. At the same time, it does not matter on what paper, what ink or symbols of what size the protocol is filled. It is important that the relationship between the results and symbols be unambiguous and the relationship between the symbols correspond to the relationship between the results of the experiment.
examples. It would be strange if, according to the protocol, where body weight is measured in grams, some conclusions were made, and according to the protocol, where body weight is measured in kilograms, others.
The experiment ends with an analysis of its results, where the hypothesis expressed in the study is confirmed or refuted. To do this, the results achieved at the end of the experiment are compared with the initial level of knowledge about the state of the subject of research.
For example, if at MPC copper 0.1 mg/l we get data on objects a, b, c... 0.2; 0.3; 0.5, it can be argued that the object is contaminated with copper cations above the MPC by 2, 3, 5 times, respectively. If the results turn out to be ambiguous, for example, when determining the MPC of copper in a qualitative way, data on objects were obtained a = 0.3 mg/l; c = 0.4 mg/l; c \u003d 0.5 mg / l, and quantitatively, respectively, 0.1; 0.2; 0.2 mg / l, then it will be difficult to draw a conclusion and the experiment must be continued by changing or improving the methodology.
An important element of the analysis of the results of the experiment is the ability of the researcher to develop scientific and practical recommendations. Recommendations should indicate clear boundaries of the possible application of the experimental system in practice.
For example, in the course of the experiment, the expediency of using X-class fertilizers in given climatic conditions, for a given type of soil, was proved to reduce the growing season of the y variety. X-fertilizers can also be recommended for varieties yx, U2, Uz. At the same time, the effect on variety Z turned out to be insignificant (or costly), and a negative result was obtained for variety F.
It is also necessary to evaluate the cost side of the experiment. If, for example, the yield of the experimental plot increased by 30% compared to the control plot, and the amount of costs increased by 1.5-2 times, then the results of the experiment are rather negative than positive, therefore, it is necessary to give balanced, cautious estimates.
So, when summing up the results of the experiment, the effectiveness of the result, its optimality with
in terms of compliance with the maximum capabilities of this system and the time spent, the conditions for the effective application of recommendations, the boundaries of successful application and restrictions under which the effect may not be optimal.
Psychological experiment- an experiment conducted in special conditions to obtain new scientific knowledge about psychology through the targeted intervention of a researcher in the life of the subject.
Various authors interpret the concept of "psychological experiment" ambiguously; often, under the experiment in psychology, a complex of different independent empirical methods is considered ( actual experiment, observation, questioning, testing). However, traditionally in experimental psychology, the experiment is considered an independent method.
Within the framework of psychological counseling, a psychological experiment is a specially created situation designed for a more holistic (in various modalities) experience by the client of his own experience.
The specifics of a psychological experiment
In psychology, experimental research has its own specifics, which makes it possible to consider it separately from research in other sciences. The specifics of the psychological experiment is that:
- The psyche as a construct cannot be directly observed and one can learn about its activity only based on its manifestations, for example, in the form of a certain behavior.
- When studying mental processes, it is considered impossible to single out any one of them, and the impact always occurs on the psyche as a whole (or, from a modern point of view, on the body as a single indivisible system).
- In experiments with humans (as well as some higher animals, such as primates), there is an active interaction between the experimenter and the subject.
- This interaction, among other things, makes it necessary for the subject to have instructions (which, obviously, is not typical for natural science experiments).
General information
In a simplified example, the independent variable can be considered as a relevant stimulus (St(r)), the strength of which is varied by the experimenter, while the dependent variable is the reaction ( R) of the subject, his psyche ( P) on the impact of that relevant stimulus.
However, as a rule, it is precisely the desired stability of all conditions, except for the independent variable, that is unattainable in a psychological experiment, since almost always, in addition to these two variables, there are also additional variables, systematic irrelevant incentives (St(1)) and random stimuli ( St(2)), leading to systematic and random errors, respectively. Thus, the final schematic representation of the experimental process looks like this:
Therefore, three types of variables can be distinguished in the experiment:
- Additional variables (or external variables)
So, the experimenter is trying to establish a functional relationship between the dependent and independent variable, which is expressed in the function R=f( St(r)), while trying to take into account the systematic error that arose as a result of exposure to irrelevant stimuli (examples of a systematic error include the phases of the moon, time of day, etc.). To reduce the likelihood of the impact of random errors on the result, the researcher seeks to conduct a series of experiments (an example of a random error can be, for example, fatigue or a mote that has fallen into the eye of the test subject).
The main task of the experimental study
The general task of psychological experiments is to establish the existence of a connection R=f( S, P) and, if possible, the form of the function f (there are various types of relationships - causal, functional, correlation, etc.). In this case, R- test subject's response S- the situation and P- the personality of the subject, the psyche, or "internal processes". That is, roughly speaking, since it is impossible to “see” mental processes, in a psychological experiment, based on the reaction of subjects to stimulation regulated by the experimenter, some conclusion is made about the psyche, mental processes or personality of the subject.
Stages of the experiment
Each experiment can be divided into the following stages. The first stage is the formulation of the problem and goal, as well as the construction of an experiment plan. The plan of the experiment should be built taking into account the accumulated knowledge and reflect the relevance of the problem.
The second stage is the actual process of active influence on the surrounding world, as a result of which objective scientific facts are accumulated. Properly selected experimental technique contributes to obtaining these facts to a large extent. As a rule, the experimental method is formed on the basis of those difficulties that must be eliminated in order to solve the problems posed in the experiment. A technique developed for some experiments may be suitable for other experiments, that is, acquire universal significance.
Validity in a psychological experiment
As in natural science experiments, so in psychological experiments, the concept of validity is considered the cornerstone: if the experiment is valid, scientists can have some confidence that they measured exactly what they wanted to measure. A lot of measures are taken in order to respect all kinds of validity. However, it is impossible to be absolutely sure that in some, even the most thoughtful, study, all the validity criteria can be completely met. A completely flawless experiment is unattainable.
Classifications of experiments
Depending on the conditions for conducting, allocate
- Laboratory experiment - the conditions are specially organized by the experimenter. The main objective is to ensure high internal validity. The allocation of a single independent variable is characteristic. The main way to control external variables is elimination (elimination). External validity is lower than in the field experiment.
- Field, or natural experiment - the experiment is carried out in conditions that the experimenter does not control. The main task is to ensure high external validity. The selection of a complex independent variable is characteristic. The main ways to control external variables are randomization (the levels of external variables in the study correspond exactly to the levels of these variables in life, that is, outside the study) and constancy (make the level of the variable the same for all participants). Internal validity is generally lower than in laboratory experiments.
Depending on the result of the impact,
Ascertaining experiment - the experimenter does not irreversibly change the properties of the participant, does not form new properties in him and does not develop those that already exist.
Formative experiment - the experimenter changes the participant irreversibly, forms in him such properties that did not exist before or develops those that already existed.
Pathopsychological experiment - the purpose of the experiment is the task of qualitative and quantitative assessment of the main processes of thinking; the experimenter, as a rule, is not interested in the immediate results of testing, since research is carried out during the experiment way achieving a result.
depending on the level of awareness
Depending on the level of awareness, experiments can also be divided into
- those in which the subject is given complete information about the goals and objectives of the study,
- those in which, for the purposes of the experiment, some information about him from the subject is withheld or distorted (for example, when it is necessary that the subject does not know about the true hypothesis of the study, he may be told a false one),
- and those in which the subject is unaware of the purpose of the experiment or even of the very fact of the experiment (for example, experiments involving children).
Organization of the experiment
Flawless Experiment
Not a single experiment in any science is able to withstand the criticism of the supporters of the "absolute" accuracy of scientific conclusions. However, as a standard of perfection, Robert Gottsdanker introduced the concept of “perfect experiment” into experimental psychology - an unattainable ideal of an experiment that fully satisfies the three criteria (ideality, infinity, full compliance), to which researchers should strive to approach.
A flawless experiment is a model of experiment that is impracticable in practice and is used as a benchmark by experimental psychologists. This term was introduced into experimental psychology by Robert Gottsdanker, the author of the well-known book "Fundamentals of Psychological Experiment", who believed that the use of such a sample for comparison would lead to a more effective improvement of experimental methods and the identification of possible errors in planning and conducting psychological experiment.
Criteria for a flawless experiment
A flawless experiment, according to Gottsdanker, must satisfy three criteria:
- Ideal experiment (only independent and dependent variables change, there is no influence of external or additional variables on it)
- Infinite experiment (the experiment must continue indefinitely, since there is always the possibility of a manifestation of a previously unknown factor)
- An experiment of full correspondence (the experimental situation must be completely identical to how it would happen "in reality")
Interaction between experimenter and subject
The problem of organizing interaction between the experimenter and the subject is considered one of the main problems generated by the specifics of psychological science. The instruction is considered as the most common means of direct communication between the experimenter and the subject.
Instruction to the subject
The instruction to the subject in a psychological experiment is given in order to increase the likelihood that the subject adequately understood the requirements of the experimenter, so it gives clear information on how the subject should behave, what he is asked to do. For all subjects within the same experiment, the same (or equivalent) text with the same requirements is given. However, due to the individuality of each subject, in experiments the psychologist is faced with the task of ensuring an adequate understanding of the instruction by the person. Examples of differences between subjects that determine the appropriateness of an individual approach:
- it is enough for some subjects to read the instruction once, for others - several times,
- some subjects are nervous, while others remain cool,
- etc.
Requirements for most instructions:
- The instruction should explain the purpose and significance of the study
- It must clearly state the content, course and details of the experience.
- It should be detailed and at the same time sufficiently concise.
Sampling problem
Another task facing the researcher is the formation of a sample. The researcher first of all needs to determine its volume (number of subjects) and composition, while the sample must be representative, that is, the researcher must be able to extend the conclusions drawn from the results of the study of this sample to the entire population from which this sample was collected. For these purposes, there are various strategies for selecting samples and forming groups of subjects. Very often, for simple (one-factor) experiments, two groups are formed - control and experimental. In some situations, it can be quite difficult to select a group of subjects without creating a selection bias.
Stages of a psychological experiment
The general model for conducting a psychological experiment meets the requirements of the scientific method. When conducting a holistic experimental study, the following stages are distinguished:
- Initial problem statement
- Statement of a psychological hypothesis
- Working with scientific literature
- Search for definitions of basic concepts
- Compilation of a bibliography on the subject of the study
- Refinement of the hypothesis and definition of variables
- Definition of experimental hypothesis
- Choice of an experimental tool that allows:
- Manage independent variable
- Register dependent variable
- Planning a Pilot Study
- Highlighting Additional Variables
- Choosing an Experimental Plan
- Formation of the sample and distribution of subjects into groups in accordance with the adopted plan
- Conducting an experiment
- Experiment preparation
- Instructing and motivating subjects
- Actually experimentation
- Primary data processing
- Tabulation
- Information Form Transformation
- Data validation
- Statistical processing
- Choice of statistical processing methods
- Converting an Experimental Hypothesis to a Statistical Hypothesis
- Carrying out statistical processing
- Interpretation of results and conclusions
- Recording the research in a scientific report, monograph, letter to the editor of a scientific journal
Advantages of the experiment as a research method
The following main advantages that the experimental method has in psychological research can be distinguished:
- Possibility to choose the start time of the event
- The frequency of the event under study
- Changeability of results through conscious manipulation of independent variables
- Ensures high accuracy of results
- Repeated studies under similar conditions are possible
Control methods
- Exclusion method (if a certain feature is known - an additional variable, then it can be excluded).
- Equalization method (used when one or another interfering feature is known, but it cannot be avoided).
- Randomization method (used if the influencing factor is not known and it is impossible to avoid its impact). A way to retest the hypothesis on different samples, in different places, on different categories of people, etc.
Criticism of the experimental method
Supporters of the unacceptability of the experimental method in psychology rely on the following provisions:
- The subject-subject relationship violates scientific rules
- The psyche has the property of spontaneity
- The mind is too fickle
- The mind is too unique
- The psyche is too complex an object of study
Psychological and pedagogical experiment
A psychological and pedagogical experiment, or a formative experiment, is a type of experiment that is specific exclusively to psychology, in which the active influence of the experimental situation on the subject should contribute to his mental development and personal growth.
A psychological and pedagogical experiment requires a very high qualification on the part of the experimenter, since the unsuccessful and incorrect use of psychological methods can lead to negative consequences for the subject.
Psychological and pedagogical experiment is one of the types psychological experiment.
In the course of a psychological and pedagogical experiment, the formation of a certain quality is supposed (that is why it is also called "forming"), usually two groups participate: experimental and control. The participants of the experimental group are offered a certain task, which (according to the experimenters) will contribute to the formation of a given quality. The control group of subjects is not given this task. At the end of the experiment, the two groups are compared with each other to evaluate the results.
The formative experiment as a method appeared thanks to the theory of activity (A.N. Leontiev, D.B. Elkonin, etc.), which affirms the idea of the primacy of activity in relation to mental development. During the formative experiment, active actions are performed by both the subjects and the experimenter. On the part of the experimenter, a high degree of intervention and control over the underlying variables is required. This distinguishes experiment from observation or examination.
natural experiment
A natural experiment, or field experiment, in psychology, is a type of experiment that is carried out under the conditions of normal life of the subject with a minimum of experimenter intervention in this process.
When conducting a field experiment, it remains possible, if ethical and organizational considerations allow, to leave the subject in the dark about his role and participation in the experiment, which has the advantage that the fact of conducting the study will not affect the natural behavior of the subject.
A laboratory experiment, or an artificial experiment, is carried out in artificially created conditions (within a scientific laboratory) and in which, as far as possible, the interaction of the studied subjects is ensured only with those factors that are of interest to the experimenter. Subjects under study are considered to be subjects or a group of subjects, and the factors of interest to the researcher are called relevant stimuli.
The specificity that distinguishes a psychological laboratory experiment from experiments in other sciences lies in the subject-subject nature of the relationship between the experimenter and the subject, which is expressed in active interaction between them.
A laboratory experiment is set up in cases where the researcher needs to provide the greatest possible control over the independent variable and additional variables. Additional variables are called irrelevant, or irrelevant, and random stimuli, which in natural conditions are much more difficult to control.
Control over additional variables
As a control over additional variables, the researcher should carry out: Finding out all irrelevant factors that can be identified If possible, keeping these factors unchanged during the experiment Tracking changes in irrelevant factors during the experiment
Pathopsychological experiment
The pathopsychological diagnostic experiment has specific differences from the traditional test research method in terms of the research procedure and analysis of the research results in terms of qualitative indicators (the absence of a time limit on the task, the study of the method for achieving the result, the possibility of using the experimenter's help, verbal and emotional reactions during the task, etc.). P.). Although the stimulus material of the techniques itself may remain classical. This is what distinguishes the pathopsychological experiment from the traditional psychological and psychometric (test) research. Analysis of the protocol of a pathopsychological study is a special technology that requires certain skills, and the "Protocol" itself is the soul of the experiment.
One of the basic principles for constructing experimental techniques aimed at studying the psyche of patients is the principle of modeling ordinary mental activity carried out by a person in work, study, and communication. Modeling consists in isolating the main mental acts and actions of a person and provoking or, better to say, organizing the performance of these actions in unusual, somewhat artificial conditions. The quantity and quality of such models are very diverse; here is analysis, and synthesis, and the establishment of various connections between objects, combination, dismemberment, etc. In practice, most experiments consist in the fact that the patient is offered to do some work, they are offered a number of practical tasks or actions "in the mind", and then they carefully record how the patient acted, and if he made a mistake, then what caused and what type of these errors were
Methodology is the total and mental and physical operations placed in a certain sequence, in accordance with which the goal of the study is achieved.
When developing methods for conducting an experiment, it is necessary to provide for:
Carrying out a preliminary targeted observation of the object or phenomenon under study in order to determine the initial data (hypotheses, selection of varying factors);
Creation of conditions in which experimentation is possible (selection of objects for experimental exposure, elimination of the influence of random factors);
Determination of measurement limits; systematic observation of the course of development of the phenomenon under study and accurate descriptions of the facts;
Carrying out systematic registration of measurements and assessments of facts by various means and methods;
Creation of repetitive situations, changing the nature of conditions and cross-effects, creation of complicated situations in order to confirm or refute previously obtained data;
The transition from empirical study to logical generalizations, to analysis and theoretical processing of the received factual material.
Before each experiment, its plan (program) is drawn up, which includes:
Purpose and objectives of the experiment;
Choice of varying factors;
Justification of the scope of the experiment, the number of experiments;
The procedure for the implementation of experiments, determining the sequence of changing factors;
Choice of the factor change step, setting intervals between future experimental points;
Justification of measuring instruments;
Description of the experiment;
Substantiation of methods for processing and analyzing the results of the experiment.
Experimental results must meet three statistical requirements:
The requirement for the effectiveness of assessments, i.e. minimal deviation variance relative to the unknown parameter;
The requirement for consistency of assessments, i.e. with an increase in the number of observations, the parameter estimate should tend to its true value;
The requirement of unbiased estimates is the absence of systematic errors in the process of calculating the parameters.
The most important problem in conducting and processing the experiment is the compatibility of these three requirements.
Elements of the theory of experiment planning
The mathematical theory of the experiment determines the conditions for the optimal conduct of the study, including in the case of incomplete knowledge of the physical essence of the phenomenon. For this, mathematical methods are used in the preparation and conduct of experiments, which makes it possible to investigate and optimize complex systems and processes, to ensure high efficiency of the experiment and the accuracy of determining the factors under study.
Experiments are usually carried out in small series according to a pre-agreed algorithm. After each small series of experiments, the results of observations are processed and a strictly justified decision is made about what to do next.
When using the methods of mathematical planning of the experiment, it is possible:
Solve various issues related to the study of complex processes and phenomena;
Conduct an experiment in order to adapt the technological process to changing optimal conditions for its flow and thus ensure high efficiency of its implementation, etc.
The theory of mathematical experiment contains a number of concepts that ensure the successful implementation of research tasks:
The concept of randomization;
The concept of sequential experiment;
Concept of mathematical modeling;
The concept of optimal use of the factor space and a number of others.
Principle of randomization lies in the fact that an element of chance is introduced into the experimental plan. To do this, the design of the experiment is drawn up in such a way that those systematic factors that are difficult to control are taken into account statistically and then excluded from the studies as systematic errors.
When carried out sequentially the experiment is not carried out simultaneously, but in stages, so that the results of each stage are analyzed and a decision is made on the advisability of further research ( fig.2.1 ). As a result of the experiment, a regression equation is obtained, which is often called a process model.
For specific cases mathematical model is created on the basis of the target orientation of the process and objectives of the study, taking into account the required accuracy of the solution and the reliability of the initial data.
An important place in the theory of experimental design is occupied by optimization issues investigated processes, properties of multicomponent systems or other objects.
As a rule, it is impossible to find such a combination of values of the influencing factors, in which the extremum of all response functions is simultaneously reached. Therefore, in most cases, only one of the state variables, the response function characterizing the process, is chosen as the optimality criterion, and the rest are accepted as acceptable for this case.
Methods for planning an experiment are currently developing rapidly, which is facilitated by the possibility of widespread use of computers.
Computational experiment called the methodology and technology of research based on the use of applied mathematics and electronic computers as a technical base when using mathematical models.
Thus, a computational experiment is based on the creation of mathematical models of the objects under study, which are formed with the help of some special mathematical structure that can reflect the properties of the object that it manifests under various experimental conditions.
However, these mathematical structures turn into models only when the elements of the structure are given a physical interpretation, when the relationship between the parameters of the mathematical structure and the experimentally determined properties of the object is established, when the characteristics of the elements of the model and the model itself as a whole find correspondence with the properties of the object.
Thus, mathematical structures, together with a description of the correspondence to the experimentally discovered properties of an object, are a model of the object under study, reflecting in a mathematical, symbolic (sign) form the dependencies, relationships and laws objectively existing in nature.
Each computational experiment is based both on a mathematical model and on the methods of computational mathematics. Modern computational mathematics consists of many sections developing along with the development of electronic computing technology.
On the basis of mathematical modeling and methods of computational mathematics, the theory and practice of a computational experiment was created, the technological cycle of which is usually divided into the following stages.
1. For the object under study, a model is built, usually first a physical one, fixing the division of all factors acting and the phenomenon under consideration into main and secondary factors, which are discarded at this stage of the study.
2. A method for calculating the formulated mathematical problem is being developed. This task is presented in the form of a set of algebraic formulas, according to which calculations and conditions should be carried out, showing the sequence of application of these formulas; the set of these formulas and conditions is called a computational algorithm.
The computational experiment has a multivariant character, since the solutions of the tasks set often depend on numerous input parameters.
In this regard, when organizing a computational experiment, one can use effective numerical methods.
3. An algorithm and a program for solving the problem on a computer are being developed. Decision programming is now determined not only by the artist's art and experience, but is developing into an independent science with its own fundamental approaches.
4. Carrying out calculations on a computer. The result is obtained in the form of some digital information, which will then need to be decrypted. The accuracy of information is determined in a computational experiment by the reliability of the model underlying the experiment, the correctness of algorithms and programs (preliminary "test" tests are carried out).
5. Processing of calculation results, their analysis and conclusions. At this stage, it may be necessary to refine the mathematical model (complication or, conversely, simplify), proposals for creating simplified engineering solutions and formulas that make it possible to obtain the necessary information in a simpler way.
A computational experiment acquires exceptional significance in those cases where full-scale experiments and the construction of a physical model turn out to be impossible.
In science and technology, many areas are known in which a computational experiment is the only possible one in the study of complex systems.