After reading this article you will learn about the quantitative and qualitative analysis for measuring behaviour of an individual.
Measurement in psychology employs two types of data or responses; verbal data and behavioural data. Verbal data refers to what a person says or writes by using language. For example, when a psychologist is trying to study a person who is depressed, he may ask the person to narrate what exactly is happening to him or write in detail about what he is undergoing.
Behavioural data, on the other hand, refers to other forms of bodily responses such as muscular, glandular, sensory, neural, perceptual etc. The same depressed individual could be studied by the psychologist by carefully observing his facial expressions, tone, posture, etc. The reader is already familiar with the distinction between verbal tests and non-verbal tests.
This use of verbal and non-verbal material is not confined only to intelligence testing but also extends to other aspects of behaviour. For instance, personality traits such as sociability, assertiveness, tolerance, love of privacy etc., can be measured through various personality questionnaires and inventories like Cattell’s Sixteen Personality Factor Questionnaire (16 PF), Eysenck’s Personality Inventory, Minnesota Multiphasic Personality Inventory (MMPI), etc. These traits can also be measured through behavioural changes such as bodily changes, cortical arousal, observation of learning behaviour, etc. In many instances both verbal and non-verbal measures are used.
We may examine the various indicators which are employed to measure verbal and non-verbal behaviour.
Some of the important indicators which are employed to measure verbal and non-verbal behaviour are mentioned below:
1. Response Time or Latency:
One factor which is normally employed to measure behaviour is the time taken for an individual to produce a response. A classical example of this is the reaction-time experiment.
2. Duration of Response:
Another factor which is taken into consideration for measuring behaviour is the duration of time for which a particular behaviour of response occurs. Measurements of after images and such other sensory experiences are subjected to this type of indices. Suppose you look at a bright green light.
The experienced greenness may remain for a moment even after you cease looking at the light. Similarly, when you hear a loud sound prolonged for sometime or inhale a strong perfume for a long time the sound and perfume will remain for sometime even after these stimuli are withdrawn.
3. Time Taken for a Response to be Completed:
This measure is used very widely in measuring learning, intelligence and other abilities. For example, in Skinner’s learning experiment or Thorndike’s trial-and-error learning experiment, one of the criteria employed to measure whether the rat or cat has learnt the correct path is in terms of the time taken by the animal to reach its goal.
4. Frequency of Response:
The number of times a particular response occurs within a given time or on a particular occasion is another indicator. An example of this type can be seen in the measurement of fluctuation of attention. Experiments on fluctuation of attention employ, as an index, the number of times attention shifts from one aspect of a given stimulus to another within a stipulated time limit.
5. Amount of Response:
In measuring emotional behaviour the amount or intensity of glandular and muscular responses is employed as an indicator. If a person’s aggression has to be measured, then the experimenter may try to measure the subject’s blood pressure, rate of respiration, rate of heartbeat, gestures, tone, facial and other expressions accompanied by certain psychological changes. Only after analysing and combining a variety of such data does one may arrive at the measure indicating overall reaction of aggression or total amount of aggressive reaction.
6. Number of Trials Required:
Yet, another indicator used is the number of trials, practice attempts or presentations of a certain stimulus. This is very commonly used in experiments on learning. In most of the learning experiments the number of attempts required by an organism to learn a task to a standard or criterion is used as an index. Similarly, experiments on remembering and learning also employ a number of presentations or trials required for a person to learn verbal material to the point of perfect recall.
7. Number of Correct Responses and Wrong Responses:
Learning experiments as well as ability tests use the total number of correct responses or wrong responses as indicators. For example, intelligence, achievement, memory and other cognitive factors are measured by various intelligence tests such as Binet’s intelligence test, and Wechsler’s intelligence test.
In all such instances the performance of a subject is finally interpreted in terms of intelligence quotient or high or low aptitude for a specific skill. This is usually arrived at by employing the criterion of assessing high scores and giving more weightage to correct responses and vice versa.
8. Response Deviation:
Yet another measured indicator is the degree or extent to which an individual response differs or varies from the normal or averages response. This is very commonly used in the measurement of abilities, personality etc. The most suitable example to understand this criterion is that IQs can be interpreted according to the norms of the normal probability curve.
Distribution of various IQs on this curve indicates whether intelligence of the subjects is below average, average or above average. This also indicates the direction of the deviation, whether it is towards the positive or negative side. In other words whether the subject is gifted, average or retarded and if so to what extent.
9. Complexity and Difficulty of Response:
The more complex and difficult a particular response, the higher the score. The concept of mental age is based on this. Some of the items of aptitude tests and intelligence tests are planned in such a way that the difficulty level is deliberately increased.
For instance, in Binet’s intelligence test the items are arranged in such a manner that the complexity and difficulty level is increased gradually as the test advances. Thus, we see that psychologists employ different kinds of measures depending on the nature of the behaviour and the purpose of the measurement.
Relative vs. Absolute Scores:
A point to be borne in mind is the distinction between relative and absolute measures. When somebody is described as five-foot tall, this means his height is five feet. This is an example of absolute measure. Wherever his height is measured it would be the same and its value will not change. On the other hand, when the same person is described as having an IQ of 110, this is not an absolute measure.
This depends on the test administered. If a test with different norms is administered, the person’s IQ may turn out to be much higher or lower than 110. Here, we see that IQ is not an absolute measure but a relative measure. The person s IQ is reported as 110 in relation to the performance of other people belonging to his age-group, i.e. it is in comparison with a norm or standard. If the norm changes the IQ will also change.
To a very large extent psychological scores or measures are relative measures. They are always arrived at with reference to certain norms unlike physical measures. In view of this, psychological measure such as IQ should be interpreted very carefully. Similarly, when a person is described as extroverted, this is again with reference to certain average or norms.
It is a limitation that psychologists have to largely depend on relative measures and not on absolute measures. But this difficulty has not prevented them from developing more and more tools and techniques measuring different aspects of behaviour.
Such measurements have also been found to be very useful in understanding and predicting behaviour. Psychological measurement is widely employed today not only in laboratories but for practical purposes such as selecting people for jobs, diagnosing psychological abnormalities and so on.
Reliability and Validity:
Whatever be the type of measures employed, it is necessary that such measures should satisfy certain criteria. This is very necessary for developing scientific laws and generalisations and also for using these measures for practical purposes.
Firstly, psychological measures like any other measures should be accurate and precise.
Secondly, they should be sensitive. For example, if a test of intelligence gives the IQ of a person as 120 then this test should be able to measure accurately any changes in the IQ.
Similarly, other measures such as attitude scales should be sensitive so that even very small changes in attitudes should be indicated by the scores. Thus, accuracy, precision and sensitivity are important requirements.
Psychological tools should be carefully constructed so that the scores or measures given by them can be depended upon. This quality is known as reliability. A particular tool, if it yields one score today and another tomorrow and a third score on another day then this score or measure is not dependable. This is why psychological scores are usually given on the basis of the average of a number of performances and rarely on the basis of a single performance.
Here we can see a difference between physical measurement and psychological measurement. Reliability is a very important characteristic and psychologists have developed different methods of ensuring the reliability of measurements.
Yet another important quality is known as validity. This means that when a psychologist claims to measure a particular psychological attribute or behaviour he should make sure that he is really measuring the particular quality such as extroversion. We must make sure that the test really measures extroversion and not something else.
For example, a thermometer measures temperature and not atmospheric pressure. Similarly, a test of intelligence should measure intelligence and not something else. This particular quality of measuring what one purports to measure is called validity. People employing psychological measurements or scores should make sure that these requirements are taken care of.
Stimulus Measures and Response Measures:
The different indicators used in psychological measurements such as intensity of response, duration of response, latency etc. All these are based on the response or the behaviour of the person and are, therefore, called response measurements.
Such measurements, however, were not the first type of measurements used in psychology. These types of measurements had to wait for the development of proper tools and tests and also the development of adequate theories and principles of psychological measurement.
Early measurements of behaviour were made in a different manner. Scientists of the nineteenth century interested in studying the relationship between physical events or stimuli, on the one hand, and psychological responses, on the other, employed other types of measurement in their experiments.
Since proper techniques were not available for measuring responses they tried to measure responses indirectly with reference to measurement of stimuli or physical events. These scientists known as psychophysicists developed a brand of psychology known as psychophysics.
They were primarily interested in answering such questions as are given below:
(a) What should be the minimum value of a stimulus of light or sound to produce a response such as seeing or hearing? This value of the stimulus was described as absolute limen or absolute threshold.
(b) What is the minimum amount of change in the stimulus that is necessary to produce a different response? If a person is blindfolded and a hundred gram weight is placed on his palm, what is the value by which the weight is to be increased or decreased so that the person just feels a change in the weight? This was known as the differential limen or differential threshold.
It may be seen from the experiments that the variable measured is the stimulus and not the response. Scientists like Weber, Fechner and others employing such procedure developed a number of psychological laws.
Subsequently, however, this type of measurement has been given up and methods of measuring the responses directly have been developed. The reader will now be in a position to appreciate the nature of psychological measurement, the type of psychological measurement, the requisites of sound psychological measurement and finally the problems involved in psychological measurement.
Today measurement in psychology has developed considerably and achieved a reasonable degree of sophistication. But still psychological measurement is not as simple as physical measurement.
It is hoped that psychologists will be able to develop more refined and sophisticated methods and techniques of quantification so that their pursuit of understanding, predicting and controlling human behaviour will be more successful.
Concept of Sampling:
A salesman who sells certain commodities very often presents his customers with some samples so that they may use them and come to a judgement about the quality of the product. Medical representatives supply doctors with sample drugs to enable them to try the same and then prescribe the medicine to their patients.
Manufacturers of soaps and perfumes also employ a similar procedure. Scientific investigation also follows a similar procedure. Metallurgists test a small quantity of ore before estimating the metallic content of that particular type of ore. Chemists and physicists also study the properties of small quantities of substances and arrive at general conclusions about them.
Psychologists also employ similar sampling procedures. For example, if you want to develop a test to measure intelligence or personality, it is not possible for us to study an individual’s performance in a large number of activities and then come to a conclusion.
We have to limit our observations to a few tasks or activities and then come to a conclusion about his intelligence or personality. We generalize about a person’s intelligence or personality. The reader can see how this involves sampling. The limited number of activities which we observe and measure constitute a sample of the total number of activities or tasks. The latter is called the universe of behaviour.
It can therefore, be seen that our ability to give a score to a person on intelligence or personality very much depends on how carefully and accurately we select a sample of behaviour. If the wrong behaviour sample is chosen then our measurement will not be valid or reliable or accurate or precise.
It is usually said that accuracy and precision of measurement depend on how truly representative our sample of behaviour is. For example, in an intelligence test the activities selected should involve all the possible factors or operations of intelligence.
On the other hand, if we include only tasks involving memory, then we will not be measuring intelligence. The reader can, therefore, see here one major source of errors. Such errors are systematic errors. The reader can now appreciate importance of validity.
The problem of sampling is involved in psychology in another way also. Psychological measures are relative measures to a large extent and have to be interpreted with the help of norms or standards of comparison. A norm or standard is an average estimate of the performance of a particular group of people. Let us imagine that we are developing a test to measure sociability.
We have to develop norms for different age groups, sex groups, rural or urban groups, etc. This is done by administering the test to a sample of people from each group. It is not possible for practical purposes to administer the test to all the boys or all the girls or all the rural subjects.
We have to select a sample of subjects who will be truly representative of the entire group. Thus, here also the problem of sampling is involved. The sample of people or subjects should be carefully selected so that they include all the different types of people falling m that broad group.
For example, if one wants to test the sociability of rural boys one must select a sample which includes poor and rich rural boys, sons of farmers, sons of landlords, sons of craftsmen and in fact all categories.
Otherwise, the norm or average established will not be correct and consequently all measurements based on these norms will also be wrong. Thus, sampling of subjects is another problem like sampling of behaviour. Most of the errors in psychological measures result from errors in behaviour sampling and errors in population sampling.
Sampling errors from both these sources result in systematic errors which are constant and in the same direction either over-estimation or under-estimation. For example, an intelligence test which is too easy will result in an over-estimation of the intelligence of all the people on all occasions.
The random errors are unstable, unpredictable and vary from occasion to occasion and from person to person. On one occasion, it may be even an over-estimation and on another it may be an under-estimation.
Similarly, in the case of one individual there may be an over-estimation and in another an under-estimation. Random errors result from lack of uniformity in the conditions under which measurements are made. If an individual’s IQ is tested when he is fresh and alert, we will get one measure of his IQ.
If the same person is tested when he is tired and reluctant, we will get another measure of the IQ. Here we can see that such differences arise because of differences in the conditions. Measurement of behaviour, therefore, should be made under optimal and carefully standardised conditions. These conditions should be maintained uniform for different people and on different occasions.
The reader can now clearly see how defective instruments, defective norms and non-ideal conditions can distort sources. To avoid such errors resulting from defects in sampling, statisticians have evolved a number of methods or procedures for selecting samples such as random sampling, stratified sampling, etc.
It is enough if the reader gets an understanding of the problems of sampling and the importance of paying utmost attention to this aspect of measurement.
In simple terms, the term probability means the likelihood of the occurrence of an event. For example, if you toss a coin, the likelihood of heads coming up is a probability. In this instance, the answer is simple.
A coin can fall with the heads or tails up and the likelihood is that heads will be on the top 50 percent of the times, if we repeatedly toss the coin. However, if we toss it up only once the probability is 1 out of 2 or 50 percent. Similarly if we throw up a dice with numbers 1 to 6, the probability of any one of the six numbers is 1/6.
Let us now go to an experiment on learning. What happens when learning takes place? The answer is that the probability of occurrence of the correct response goes on increasing until we reach a stage of perfect learning where the learn response is almost sure to occur. In general, the nature of many aspects of human behaviour is that one cannot be totally certain that a particular response will always occur in-spite of perfect learning.
The probability of occurrence of any behaviour, when we refer to behaviour that is not unlearnt, ranges between 0 to 1. Thus in the learning experiments conducted by Thorndike and Skinner the correct response occurred by chance on the first occasion and gradually the probability of occurrence increased.
Similarly in treatment of behavioural problems with effective treatment the probability of occurrence of the problematic behaviour gradually increases. So far, we were discussing what is called ‘equal probability’. Thus, when we toss up a coin the probability of getting heads or tails is equal and similar is the case with the 6 numbers on the dice.
But in many instances of human behaviour this is not the case. For example, let us assume a psychologist tests the intelligence of a large number of people and wants to know how many may fall in different ranges of I.Q. Let us suppose he tests 100 people at random. He will find that the number of people in each range differ from the others.
If we assume that the average intelligence score is 100, then he will find that there are many more people between 90-110, and the number keeps on decreasing as we proceed away from 100. There will be very few people between 160-180.
Similarly, in range between 20 and 40. This type of distribution where one finds more number of cases in the middle region and this number progressively decreases as we proceed towards the extreme is called Gaussier Curve, a binomial distribution and most commonly normal distribution.
It has been found that a large number of human behaviour patterns follow this pattern. It is seen that most unlearnt forms of behaviour which are determined by a large number of factors follow the normal distribution. A classical example is intelligence. One doesn’t know how many factors determine the level of intelligence and we assume these factors to be genes which are either present or absent.
It is obvious that over the years psychologists have been able to develop highly precise and accurate techniques and tools of measurements. In fact, among the social and behavioural sciences psychology has emerged as the most sophisticated discipline from the point of view of measurement.
Nevertheless, the quantitative approach has its limitations. Even today, perhaps, we have not reached the stage where all aspects of behaviour can be measured quantitatively. As a result, very often qualitative analysis becomes necessary. For example, in areas like diagnosis and treatment of behavioural problems qualitative analysis has an important role to play.
The experienced clinical psychologist or therapist while using quantitative indices based on psychological tests still does not depend entirely on these for his assessment and diagnosis. Similarly, in the entire area of personality study it has been found that quantitative analysis and description are not always adequate.
Two individuals securing identical scores on a psychological test may be similar but are not identical. The fact that there are unique characteristics poses certain limitations on the validity and usefulness of quantitative analysis.
Yet another dilemma relates to the choice between the wholistic approach and the analystic approach. A number of psychologists emphasise the view that personality is a totality and has to be understood as a whole.
This approach is, to a large extent, in contrast to psychometric approach or the measurement approach. The measurement approach is essentially analytical. For example, intelligence of a person is the sum total of the scores of all the components.
Similar is the case with the personality measurement. The wholistic theories are opposed to such a view. According to them the intelligence or the personality of an individual has a wholistic or total characteristic which is missed when measurement is made in terms of units or components.
These psychologists believe that while these scores may be helpful to some extent in understanding and predicting behaviour, nevertheless, we should also keep in mind the unique and individualistic elements which ordinarily cannot be measured. Thus, one can see a disagreement between the clinical-wholistic approach on the one hand and the psychometric approach, on the other.
While the above limitations of measurement and quantitative analysis are undoubtedly valid to a certain extent, nevertheless, it should be pointed out that behaviour measures have their own usefulness and have contributed much to the development of psychology as a science. They have also helped to develop psychology as a discipline with practical applications. It would probably have been impossible to develop theories and laws of psychology without quantitative analysis and measurement.
At the same time, one should also be aware that they are not fool-proof and that there are limitations in the use of such measurement techniques. Of course, attempts are continuously being made to devise more precise, reliable and valid techniques and tools of measurement. But then there are inherent limitations arising out of the very nature of human behaviour.
This is why there is an increasing realisation that one should be more cautious in accepting quantitative measures of behaviour at their face values and many factors have to be taken into consideration while interpreting them. In general one may say that psychological test scores are to be taken in their spirit rather than the letter. While they certainly give us an indication of where the individual is on particular behaviour, they cannot be taken to be absolutely correct and much less exact.