Measurement in social science research
Download as PPTX, PDF9 likes9,836 views
1. There are four main scales of measurement in social science research: nominal, ordinal, interval, and ratio scales. 2. Nominal scales classify variables into categories with no quantitative meaning or order, while ordinal scales classify variables into ordered categories where the distances between categories are unknown. 3. Interval and ratio scales both involve ordered categories with equal distances, but a ratio scale has a true absolute zero point. 4. For a measurement to be valid and reliable, it must accurately measure the intended construct, produce consistent results over time, and be practical to administer and interpret.
Measurement in social science research
- 1. MEASUREMENT TECHNIQUES IN SOCIAL SCIENCES RESEARCH SONDARVA YAGNESH M MSc Agricultural Extension Education BACA AAU, Anand
- 2. Measurement Scales Four kinds of scale of measurement are important for quantifying variables in the behavioral sciences: 1. Nominal Scale 2. Ordinal Scale 3. Interval Scale 4. Ratio Scale
- 3. Nominal Scale ○ This type of scale allows a researcher to classify characteristics of the persons, places or objects into categories. ○ It is simply a system of assigning of number symbols to events in order to label them. ○ Example: Assignment of numbers to basket ball players to identify them and as such , the numbers have no quantitative value. ○ Sometimes variables measured on nominal scales are called categorical or qualitative. Examples: Group membership (1 = Experimental, 2=Placebo ) A person’s gender (0 = Female, 1 = Male) Blood type, marital status, religion
- 4. Nominal Scale (contd.) ○ The weakest or least powerful level of measurement ○ Indicates no order or distance relationship and has no arithmetic origin ○ Simply describes differences between things by assigning them into categories ○ Counting of numbers in each group is the only possible arithmetic operation ○ Mode as the measure of central tendency can only be used ○ Chi-square test of statistical significance can be utilized ○ Contingency coefficient for measures of correlation can be worked out.
- 5. Ordinal Scale ○ In this case, the characteristics can be put into categories and the categories also can be ordered in some meaningful way. The distance between the categories, however, is unknown. ○ A student’s rank in his class involves use of this scale. ○ Permits the ranking of items from highest to lowest but the real difference between adjacent ranks may not be equal. ○ Implies a statement of ‘greater than’ or ‘less than’ without our being able to state how much greater or less. ○ Median can be used as the measure of central tendency. ○ Percentile or quartile measure is used for measuring dispersion ○ Correlations are restricted to various rank order methods ○ Measures of statistical significance are restricted to non- parametric methods.
- 6. Ordinal Scale, Continued ○ Examples: Socioeconomic Status 1 = Low 2 = Middle 3 = High Health Status 1 = Poor 2 = Fair 3 = Good 4 = Excellent
- 7. 3. Interval Scale ❖ Numbers are assigned to objects or events which can be categorized, ordered and assumed to have an equal distance between scale values. ❖ It has an arbitrary zero, but it lacks true zero or absolute zero. ❖ It dose not have the capacity to measure the complete absence of a trait or characteristic. ❖ Example: Fahrenheit or centigrade scale of temperature ❖ Addition and subtraction are permissible, but not multiplication and division
- 8. 3. Interval Scale, Continued ❖ More powerful measurement than ordinal scale as it involves the concept of equality of interval. ❖ Mean-appropriate measure of central tendency, std. deviation most widely used measure of dispersion ❖ Product moment correlation technique ❖ ‘t’ test and ‘z’ test for statistical test of significance
- 9. 4. Ratio Scale ○ The most precise level of measurement consists of meaningfully ordered characteristics with equal intervals between them and the presence of a zero point that is not arbitrary but determined by nature. ○ For example, the zero point on a centimeter scale indicates complete absence of length or height, but absolute zero of temperature is theoretically unobtainable. ○ Represents the actual amount of variables ○ Ratio is possible, e.g. it can be said that 40 kg. is four times more than 10 kg. ○ Examples: weight, height, income, distance etc. ○ All statistical techniques are usable.
- 10. Examples of Appropriate comparison statements A is equal to (not equal to) B = (≠) A is greater than (less than) B > (<) A is three more than (less than) B + (–) A is twice (half) as large as B × (/) Relevant level of measurement NominalOrdinal Interval Ratio √ √ √ √ √ √ √ √ √ √ The Types of Comparisons That Can Be Made With Different Levels of Measurement © Pine Forge Press, an imprint of Sage Publications, 2004
- 12. Sources of error in measurement a. Respondent: ● Reluctance ● Fatigue, boredom, anxiety etc. b. Situation: c. Measurer: ● Behaviour, style or look may encourage/discourage certain replies from respondents ● Incorrect coding ● Careless mechanical processing of data ● Faulty tabulation and/or statistical calculation etc. d. Instrument: ● complex words, ambiguous meaning, poor printing, inadequate space for replies etc.
- 13. Tests of sound measurement Tests of sound measurement must meet the tests of: ❖ Validity ❖ Reliability and ❖ Practicability
- 14. ○ Measurement is said to be reliable when it give consistent results. i.e. when repeated measurements of same things give constant results. ○ Reliability is the extent to which the same finding will be obtained if the research is repeated at another time by another researcher. If the same finding can be obtained again, the instrument is consistent or reliable. ○ Reliability refers to the consistency of scores obtained by the same individuals when reexamined with test on different occasions, or with different sets of equivalent items, or under variable examining conditions. Reliability
- 16. Methods of estimating reliability coefficient ❖ Test-retest method: ➢ Single form of test is administered twice on the same sample with a reasonable time gap. ➢ It yields two independent sets of scores and the correlation between them gives the value of reliability coefficient which is also known as temporal stability coefficient.
- 17. Methods of estimating reliability coefficient ❖ Split-half method: ➢ It indicates homogeneity of the test. ➢ Test is divided into two halves, say, one set contains odd numbered items and another contains even numbered items. ➢ A single administration of the two sets of items to a sample of respondents yields two sets of scores. A positive and significant correlation indicates that the test is reliable. ➢ The advantage is that data necessary for computation of the reliability coefficient are obtained in a single administration of the test, and hence variability produced by two administrations is automatically eliminated.
- 18. Validity of measurement ○ Validity of the measuring instrument is the degree or the extent to which it measures what it is supposed to measure. ○ The term validity means truth or fidelity. It can be defined as the accuracy with which it measures that which is intended to measure. ○ Validity is epitomized by the question: ‘Are we measuring what we think we are measuring?’ This is very difficult to assess. The following questions are typical of those asked to assess validity issues: ➢ Has the researcher gained the full access to the knowledge and meanings of informants? ➢ Would experienced researcher use the same questions or methods?
- 19. ○ A good measure must not only be reliable, but also valid. ○ A valid measure measures what it is intended to measure. ○ Validity is not a property of a measure, but an indication of the extent to which an assessment measures a particular construct in a particular context—thus a measure may be valid for one purpose but not another. ○ A measure cannot be valid unless it is reliable, but a reliable measure may not be valid
- 20. Content validity ○ When the content of items individually and as a whole are relevant to the test, it represents content validity. ○ It requires both: ● Item validity: concerned with whether the test items represent measurement in the contended area, and ● Sampling validity: concerned with the extent to which the test samples the total content area.
- 21. Concurrent validity ○ In this method, a test is correlated with a criterion which is available at present time. ○ It means how well performance on a test estimates current performance on some valued measure (criterion). ○ e.g. test of dictionary skills can estimate students’ current skills in the actual use of dictionary – observation. ○ e.g. the Scholastic Aptitude Test (SAT) is valid to the extent that it distinguishes between students that do well in college versus those that do not.
- 22. Predictive validity ○ It is the degree to which a measure predicts a second future measure. ○ A test is correlated against the criterion to be made available sometimes in future. ○ Predictive Criterion Validity = how well performance on a test predicts future performance on some valued measure (criterion)? ○ e.g. reading readiness test might be used to predict students’ achievement in reading. ○ Predictive validity is needed for tests which include long range forecast of academic achievement, industrial management etc.
- 23. Construct validity ○ It is the extent to which the test may be said to measure a theoretical construct or trait. ○ A construct is a non-observable trait such as intelligence, motivation etc. ○ Construct validation is a more complex and difficult process than content validation and criterion validation. ○ Construct validity is computed only when the scope for investigating criterion related validity or content validity is bleak.
- 24. Practicability ○ From the operational point of view, the measuring instrument ought have: ❖ Economy, ❖ Convenience and ❖ Interpretability