home
about the project
Medical news
For authors
Licensed books on medicine
<< Previous Next >>

STATISTICAL PARAMETERS OF PSYCHODIAGNOSTIC RESEARCH

Each psychological sign, as a result of measurement, can be characterized by several parameters.

ACCURACY (reliability) of measurement - quality reflecting the proximity of their result to the true value of the measured quantity. High accuracy corresponds to small systematic and random errors.

CORRECTNESS OF MEASUREMENT - quality reflecting the closeness to zero of the systematic errors of its results.

CONVERGENCE of measurement - quality, reflecting the closeness to each other of the results of measurements made under the same conditions.

REPRODUCIBILITY of measurements - quality, reflecting the closeness to each other of the results of measurements performed under different conditions. In a broad sense, reproducibility depends on the stability of the characteristics of the subject.

A well-constructed test provides information that has: reproducibility (tested for reliability), derivability (allows you to obtain objective results), verifiability (allows you to use special procedures to check the validity and predictive value), the ability to predict and generalize, the ability to quantify the results obtained, practical usefulness , criticality (the ability to develop and create new modifications of specific techniques).

Obviously, accuracy (reliability) means both correctness and convergence, but converging and reproducible measurement results may be inaccurate. The value obtained during measurement and the assessment of a psychological sign are influenced by a variety of random factors from negligence in preparing stimulus material to changing solar activity. These factors usually cannot be taken into account, and the results of their influence are called random errors. The smaller the discrepancy between the measurement results under different conditions, the higher the reproducibility. To assess the quality of measurement, it is necessary to calculate the average value X of the obtained repeated results

where, X1, .... Xn are the values ​​of repeated n measurements. The smaller the difference in means, the less systematic errors and higher correctness.

The discrepancies are described by the standard deviation (standard deviation denoted by? Or SD).

To assess the accuracy of the study in two dimensions or two tests, it is necessary to evaluate the reliability of the difference between the calculated average values. To do this, calculate the average error, also called the standard error, m.

The average error indicates the accuracy with which the average value is determined, therefore, it should be calculated only for these purposes and used only for comparing average values.

If any quantitative parameter of a psychological variable is described, then the average value and standard deviation (X ??) should be given.

To assess the reproducibility of the measurement results of two tests, it is necessary to calculate the F-criterion,

where? 2 is the variance estimate of the two tests, with? 1>? 2. Dispersion is the squared standard deviation.
The value of F is compared with tabular critical values. Table 2.

The minimum value of the F criterion determining the significance of differences in variances at p <0.01.



This table is usually sufficient for comparing the convergence and reproducibility of measurements.

The coefficient of variation is a relative value and varies from 1 to 100%. If the coefficient of variation is greater than 100%, this means that the standard deviation is greater than the average value and, therefore, the distribution of the results differs from the normal (Gaussian - Laplace) and it is incorrect to describe it with the mean and standard deviation.

The result of a single measurement in one person is determined by the true value of the value of the psychological sign and the described errors called analytical variations. In turn, the true value can change for each person during the day, week, month, year, during exercise, changes in the emotional and functional state - these are internal variations.

If a researcher of people in the same conditions, then the differences between them are revealed. These differences are called interindividual variations.

In order to be able to interpret the results of the survey, taking into account the variations that affect the qualitative characteristics of the results of the survey, it is necessary to examine a representative group of people several times at different time intervals, and some measurements to be carried out repeatedly, in parallel to analyze several parts of one test. The totality of all the results obtained is defined as the total variation. The discrepancy between the results of multiple measurements on the same sample is analytical variability.

Inside, individual variability can be estimated as the difference in the variability of the results when repeated measurements in one subject and analytical variability. Interindividual variability is calculated as the difference between the total variability and the sum of the analytical and within the individual variability. It is usually understood that analytical variability should not be more than 10% of the average value of the measured characteristic.

To increase the diagnostic effectiveness of the test, it is necessary that not only the analytical variations are much smaller than the interindividual ones, but also that the individual variations inside are minimized. Otherwise, changes caused by uncontrolled factors will overshadow the differences, depending on the psychological or psychophysiological quality or property.

To achieve a decrease in the influence within individual variations on the result of a diagnostic study, it is necessary to standardize the conditions of the examination.

If studies are carried out repeatedly in the process of any influences, for example, during functional stress tests, when studying the daily or weekly working cycle of specialists, it is necessary that the analytical variations are less than individual ones, i.e., compared with the shifts that the psychophysiologist is expecting. Otherwise, the differences in the values ​​during repeated measurements will be caused to a greater extent by errors.
<< Previous Next >>
= Skip to textbook content =

STATISTICAL PARAMETERS OF PSYCHODIAGNOSTIC RESEARCH

  1. STATISTICAL STUDY METHOD
    LESSON PURPOSE: To familiarize students with the sequence of conducting statistical research, to teach them to independently plan and program statistical research, to correctly determine the unit of observation and accounting features, to familiarize themselves with the methodology for collecting material for its processing and analysis. LESSON METHOD At the beginning of the lesson, the teacher gives a definition
  2. The main psychodiagnostic method in the study of gender differences
    A survey method for identifying strategies for self-presentation of a person in clothing. This interrogation method was developed by E. A. Petrova. It consists of 13 questions with ready-made answers. The questions relate to the most common life processes that relate to fashion. The person is invited to answer questions that characterize the usual way of dressing in different situations. In interpretation
  3. NON-PARAMETRIC METHODS FOR ASSESSING THE RELIABILITY OF RESULTS OF A STATISTICAL STUDY. COMPLIANCE CRITERION (CHI-SQUARE)
    LESSON PURPOSE: To master the methodology for calculating and using the conformity criterion to determine the significance of the discrepancy between several compared groups of studied phenomena. LESSONS: The students prepare for the practical lesson on the recommended literature and complete their individual homework. The teacher checks for 10 minutes
  4. Research methods, parameters, diagnostics
    There are two main methods for determining CBS - the balancing method and the PaCO2 electrode method, which provides for direct measurement of PaCO2. Common to both methods is that in the blood sample two parameters of CBS are determined - pH and PaCO2, and the remaining parameters are calculated using Zipaarts-Andersen nomograms. The most convenient is the linearized version of the nomogram, on which
  5. REQUIREMENTS FOR THE INTERNAL CONTENT OF PSYCHEDIAGNOSTIC METHODS
    The test method of psychological diagnostics, using, as a rule, standardized questions and tasks that determine the scale of measurements. The main feature of psychodiagnostic techniques is its measuring and testing orientation, which allows for a quantitative and qualitative assessment of the studied mental properties. The test is a measuring tool for assessing one
  6. Classification of psychodiagnostic methods, their general characteristics and psychological justification
    Different authors offer different ways to classify psychodiagnostic methods. For example, according to R. S. Nemov (1995), one can distinguish: 1. Methods of psychodiagnostics based on observation (standardized and free observation); 2. Interview methods, psychodiagnostic methods (questionnaires; questionnaires, which
  7. The specifics of psychodiagnostic and psychocorrectional work with families
    Both psychodiagnostics and, subsequently, psychocorrection of family difficulties are carried out, as a rule, in the process of family counseling and family psychotherapy. The specific features of such work are determined by specific requests coming from one of the family members or the family as a whole. Specific features of psychological assistance to families are the rejection of the concept of the disease,
  8. STATISTICAL BASES OF CLASSICAL TESTING
    Currently, we can talk about the statistical foundations of classical testing and testing using new information technologies, in particular expert systems. Conventionally, we call this acmeological testing. The methodological, theoretical and methodological basis of such testing is based on the concept of factor-cumulative causes. This chapter outlines
  9. METHODS BASED ON THE STATISTICAL TRANSFORMATIONS
    STATISTICAL ANALYSIS Statistical methods are used to quantify HRV in the studied period of time. When using them, a cardiointervalogram is considered as a set of consecutive time intervals (numerical values ​​of the duration of R – R intervals). The main indicators of statistical analysis: Khsr. (Mathematical expectation, M, Меаn, RRNN) -
  10. Some statistics
    The problem of disability is one of the central places among medical and social problems. Disabled people make up about 10% of the world's population, of which 20-25% are children under the age of 16. In most countries, at least 1 out of 10 people has physical, mental or sensory defects (I.A. Skvortsov, T.N. Osipenko, 1994). Child disability rate is 4.3 per 1000 children
  11. ASSESSMENT OF STATISTICAL INDICATORS
    The most important indicators are heart rate,?, Max (s), Min (s),? X, V (%). Typically, the heart rate ranges from 60 to 90 in 1 minute. With high-quality filtering of the record, the difference between Max (s) and Min (s) - the indicator? X corresponds to the amplitude of the natural fluctuations of the rhythmogram and usually decreases with a deterioration in the functional state of the CCC. In the presence of at least one arrhythmic episode (for example,
  12. STATISTICAL BASES OF SEARCH FOR RELIABILITY AND VALIDITY OF TEST AND TEST BATTERIES
    The test method consists in the diagnosis (psycho-prognostics) of the personality, its mental states, functions, qualities based on the performance of some standardized task with a predetermined validity and reliability. We have already considered the concepts of reliability and validity. Let us dwell in more detail on the analysis of the validity of the test, the statistical basis of its calculation.
  13. STATISTICAL BASES OF SCALING IN TESTOLOGY
    The creation of working, effective tests is a complex and lengthy work requiring both a subtle understanding of the actual psychological problems of diagnosis and knowledge of the basics of mathematical statistics. Perhaps the simplest, but one of the key elements of mathematical statistics in testology is the search for the arithmetic mean (mathematical expectation) of questions and scales
  14. STATISTICAL ANALYSIS OF THE DIAGNOSTIC FORCE OF THE TEST
    STATISTICAL ANALYSIS OF DIAGNOSTIC FORCE
  15. Mathematical and statistical analysis of the results
    Based on the data obtained, it is possible to construct schemes of significant relationships between the obtained indicators. 1st group {foto14} Fig. 27 The structure of the hierarchy between significant indicators (greater than 0.75) that affect the aggressiveness of group 1 {foto15} Fig. 28 Based on the data obtained, it is possible to construct schemes of significant connections between the obtained indicators. 2 group
  16. Basic definitions and statistical concepts characterizing the antenatal, perinatal and neonatal periods of life
    A neonatologist needs to know international definitions and use them in everyday practice. The following criteria for the use of a number of concepts (such as prematurity, hyphenation, etc.) ensure the comparability of the results of the work and the reliability of the comparative analysis. Knowledge of statistical terms and methods for calculating statistical values ​​is necessary for analyzing the results of work
  17. Parametritis
    Inflammation of the perinatal fiber. The reason for the development of parametritis is most often the previous intrauterine interventions: childbirth, abortion, supravaginal amputation of the uterus, removal of the appendages, diathermocoagulation. Postpartum parametritis most often occurs with ruptures of the cervix and the upper third of the vagina. Infection with parametritis spreads lymphogenously. Diagnostics
  18. STATISTICAL DATA ON THE BASIC GROUPS OF PATIENTS AFTER INTENSIVE REHABILITATION
    During the period from 1987 to 1999, 87 patients with injuries of the spine and spinal cord of various localization underwent intensive rehabilitation using the described technology. These were mainly patients of working age. In the acute period, 7-14 days after the injury, there were 7 patients (8%) aged 43 to 52 years. Against the background of the main treatment, gentle methods of intensive
  19. STATISTICAL BASES OF EXPERT SYSTEMS IN PSYCHOLOGY, BUILT ON THE EFFECT OF CUMMULATIVE FACTOR REASONS
    Modern tests work well in the diagnosis of individual psychological characteristics of a person, however, they are not so effective when it comes to diagnosing a person’s professional suitability, and psycho-predicting their success in one or another type of activity. These shortcomings of many modern tests experienced acmeology. All this forces not just to take the achievements of modern testology, but
Medical portal "MedguideBook" © 2014-2019
info@medicine-guidebook.com