Validity of measurement results - Marketing research

Validity of measurement results

In contrast to the measurement reliability index, which reflects the repeatability of its results, the validity ( validity ) metric is used. The results of scaling are valid if the differences between the measurements revealed in a survey of different respondents are explained by the actual differences between them, rather than by a systematic or accidental measurement error. Continuing our analogy, the results of firing in a dash can be considered valid if the shooter shoots at the center of the target. Thus, reliability is a necessary condition for validity; if the integral scale is unreliable, then it certainly does not have validity.

Consider alternately three existing validation validation methods.

Content validity. Checking the content validity ( Content validity or, sometimes, Face validity ) means thinking carefully about all important aspects, which should reflect the integral scale, are contained in a set of private scales. So, if the integral scale should describe the image of the store, then you need to take care that in private scales such aspects of it as the quality of goods, variety of assortment, etc. are reflected.

Given the subjective nature of this method, it is desirable to use it only in combination with others.

Criterion validity (competitive and predictive). Using criterion validity ( Criterion validity ), the consistency of the differences in scale scores among different respondents with differences between these respondents is checked socio-demographic, psychological and behavioral characteristics. These characteristics are determined based on other questions asked by the same respondents or in the same survey ( competitive validity - concurrent validity ), or in another poll ( predictive validity - predictive validity ).

An example of verification of competitive validity can be the inclusion in the same questionnaire of two sets of private scales aimed at constructing the same integral scale; with one set - full, and the second - shortened and slightly modified. The presence of competitive validity means that the estimates thus obtained are generally consistent with each other.

An example of checking predictive validity is a situation in which the ratio of each participant of a consumer panel to a certain commodity is measured by a survey first, and then the obtained data is compared with the actual purchases of these people.

Constructional validity (convergent, discriminant and nomological). Evaluating constructional validity ( construct validity ), the researcher compares the values ​​of a given scale and other scales constructed according to the same questionnaire with the marketing theory, seeks theoretical explanations for the observed patterns, finds out why the values ​​of these scales are combined with each other in this way, and not otherwise.

The following conditions are checked. The presence of convergent ( convergent ) means that the correlation coefficients of the integral scale are quite high with other scales that measure roughly the same, but in a different way. The presence of discriminant ( discriminant ) validity means that the coefficients of correlation of the integral scale with other scales are low, with which there should not be any connection from meaningful considerations. Finally, the presence of nomologic ( nomological ) validity means that the correlation coefficients of the integral scale are large enough in modulus and have the required sign, with scales reflecting the other sides of the studied objects, with which they do not less from the substantive considerations there should be a connection. Therefore, the verification of the nomological validity of all constructed integral scales starts with the theoretical construction of the so-called nomological net ( nomological net ), on which the scales are depicted and the arrows indicate which of them should be interrelated and how (positive or negative ) way.

Let's explain the essence of the verification of structural validity in an example in which self-esteem and other characteristics of participants in this survey were studied on the basis of a survey and other information. In this paper, the following facts were noted:

o the evidence of convergent validity was that the self-assessment vector of the participants in the study is strongly and positively correlated with the vector of assessments that each of these people gave to his friends;

o the evidence of discriminant validity was that the self-assessment vector is weakly correlated with the brand loyalty rating vectors ( brand loyalty ) and the variety of brands in question when choosing a product ( variety seeking );

o evidence of nomologic validity was that respondents were more likely to prefer those brands of products whose image is closer to self-esteem in their eyes.

In addition, the researchers were convinced of the high reliability of the scales.

Generalization possibilities. The problem of external validity, or in another way - the possibility of generalizations ( generalizability ), i.e. the admissibility of the dissemination of the developed scales to other cases (survey methods, studied sets, time periods, etc.) should be decided each time separately, there are no general recommendations here. For example, to test the possibility of using the scale at other times, you need to test its reliability by the test-retest method. To test other variants of generalization, we use other methods that we will not discuss here.

Also We Can Offer!

Other services that we offer

If you don’t see the necessary subject, paper type, or topic in our list of available services and examples, don’t worry! We have a number of other academic disciplines to suit the needs of anyone who visits this website looking for help.

How to ...

We made your life easier with putting together a big number of articles and guidelines on how to plan and write different types of assignments (Essay, Research Paper, Dissertation etc)