Term Definition Essay

Introduction

Statistics is a science that deals with making inferences on data. It is divided into two descriptive statistics and inferential statistics. Descriptive statistics deals with data.

It relates to data collection, data summary, data assembly, data organizing and graphically displaying data whereas inferential statistics involves making inferences, testing for hypothesis, determination of relationships and making predictions.

Define and discuss each of the following terms in a short essay

Validity

The word validity can be used interchangeably with the words authentic, genuine, real, and truthful to mean level or degree that something is true. The following are definitions of the word as used in different fields:

Mathematics: the word valid in mathematics means correctness, the degree of which a figure is closest to the desired results.

1 / 7

Term Definition Essay

In research; the word is mainly used in statistics whereby a sample predicts or measures accurately.

In Law: the word valid is used when defining an agreement, an offer, and a claim that remains in force.

In Banking: it refers to the period that the letter of credit remains effective and has met all the requirements. (Zeller, 1979).

Content validity

Synonyms of the word content validity are logical and rational validity. Content validity is a non-statistical and qualitative type of validity that measures the extent to which a measure represents every single specific detail of its construct. A test could only be content valid if it is build carefully with items that are supposed to be included in it, it shows that elements are relevant to the construct and are free from bias. It answers the question; does the element detract from the research program?

Predictive validity

Predictive validity refers to tests that can be quantified by the correlation coefficient which is symbolized as “r”. For example if correlation coefficient is equal to 95 it means the variables being compared are 95% compatible.

2 / 7

Term Definition Essay

The predictive validity is mainly used for aptitude tests before employment to predict a certain quality that can be measured at some time in the future. It shares similarities with the concurrent validity since it measures agreement two types of measurement instruments.

Concurrent validity

Concurrent validity is a type of criterion related validity used in behavioral sciences such as sociology, and psychology. It demonstrates where a test correlates with a high percentage with a measure previously validated. For example, motivation could be correlated with work performance both measures are taken at the same time.

The difference between concurrent validity and predictive validity is in concurrent measures are done at the same time and for predictive one measure is done before the other.

Construct validity

The concept of construct validity is used in social sciences, psychology and education, to determine if a concept matches a specific measurement or scale. It refers to whether a scale or test measures the construct adequately.

3 / 7

Term Definition Essay

Construct validity seeks agreement between a theory and specific devices or procedures used for measuring. (Mesick, 1995).

Construct validity is divided into two categories: Convergent validity is the actual relationship ratings, tests or measures gathered independently. Discriminate validity is the lack of a relationship among measures which theoretically are not supposed to be related.

For construct validity to be understood, the following steps are important; First, specification of the theoretical relationships then the examination of the empirical relationships between measures and finally interpretation of the empirical evidence.

Reliability

Means the ability of an item, to consistently perform its intended function on demand or within a stipulated period of time and under specific conditions, without failure. Reliability shows freedom from error and potential bias.

Reliability could also mean the ability of a test to provide the same kind of results each time the same variables are used. It is difficult to test reliability effectively but the following tests can be undertaken as tests for reliability; Inter-rater or inter-observer Reliability, Test-Retest, Parallel-forms and Internal consistency reliability. (Zeller, 1979).

Test Retest Reliability

4 / 7

Term Definition Essay

This is a test for reliability and just as the name suggests, it involves two different kinds of tests. The same test is done on the same sample but on two different occasions. This test relies on the assumption that the sample remains the same on the stipulated period of time. The time gap between the two tests affects the correlation between the variables. The shorter the period the higher the correlation and the vice versa is also true. Therefore, the interval between tests is very crucial and should be taken into consideration. (Zeller, 1979).

Pragmatism

Pragmatism is a philosophical term that means relying on practical thinking rather than the use of theories and principles. Meanings of theories should be taken on the practicability of the concept and the underlying truth. For believes to pass as true, they have to be verified, but some concepts vary as truths depending on different circumstances. For a statement to be classified as a truth it is supposed to contain a concept that cannot be lost with time, and a concept that applies for everyone that is what is true for me is true for you. (Mesick, 1995).

Split-half method

This is a test for reliability where items used for measure are divided randomly maybe into two. The same procedures are carried out for the two sets of measure from the same sample and a correlation for each of the two are calculated and compared. The computer could be used to randomly select the values that are going to be used and calculate the correlation coefficient for each for each set of data. The following steps should be undertaken when doing the spit-half method; Divide the tests into two halves and create the scores, compute the differences then compute the variances and finally the reliability estimate. (Zeller, 1979).

Factorial validity

5 / 7

Term Definition Essay

This is a type of construct validity established through factor analysis; which involves a set of mathematical procedures to come up with the underlying interrelationships between variables. The test is carried out to determine the strength of relation between variables. Examples of test carried out for the factor validity is the chi-square test which answers the following; is the data from normal distribution, exponential distribution, logistic distribution, binomial distribution or log-normal distribution. (Mesick, 1995).

6 / 7

Term Definition Essay

References

Edward G. Carmines and Richard A. Zeller (1979). Reliability and Validity Assessment:

Quantitative Applications in the Social Sciences. Sage Publishers.

Mesick, S. (1995). Validity of psychological assessment: Validation of inferences from persons’ responses and performances as scientific inquiry into score meaning. American Psychologist.