Social science researchers have a responsibility towards effective assessment and measurement of works conducted by others through appropriate usage of test and scaling tools (Sahn & Stifel, 2000, p.96). One, therefore, as a social science researcher needs to acquaint himself or herself with these tools to be able to carry out measurement of data in several disciplines as may be required (Krzanowski & Marriott, 1994a). For most of the time, a critique of the tools used by others becomes very necessary for self-development (Chatfield & Collins, 1980). A scaling method, by explanation, has to do with organizing data in terms of quantitative attributes (Everitt & Dunn, 2001, p.18; Everitt et. al, 2001, p.68). Lately, easier-to-use scaling tools such as the SPSS have been developed to institute a change in how scientific data is been dealt with. With such a tool, analysts are better equipped to transform variables by simply clicking a bottom (Oppenheim, 1992). Even though so much has been put together to enhance data analysis, several times, statistical tasks have been considered arduous by researchers. There is always an emergent obstacle that researchers will find distressing rather than been exciting and presenting mastery for important skills.
A number of other scale and test tools, including Wechsler memory- used in the estimation of the memory function of an individual, MMPI- used in the measurement of mental alertness, truthfulness scale- used in determining how reliable one’s truth is, and the Bayley for determining a child’s proper growth-rate, among others, have been put in place for accessing variables (Manly, 1994, p.24).
One other scaling method that is found very useful in social science is the General Linear Model (or the GLM) which is a tool in statistics that is used to effectively incorporate the normal distribution of dependent variables as well as variables that are continuously independent (GenStat, 2002). The GLM’s SPSS procedures affords one the opportunity to operate with specifically generalized linear models in terms of syntactic or dialoguing boxes, and equally makes it possible for one to get outcomes in pivoted tabular format- this is of considerable significance based on the fact that the GLM makes easy the editing of outcomes (De Vaus, 1990, p.36; Krzanowski & Marriott, 1994b, p.6). Otherwise, the several features present on the GLM, therefore, make it possible and easier for one to put together designs that have vacant cells by plotting mean estimations and customizing linearly structured models in conformity with an available research question (Sahn & Stifel, 2000, p.92). Researchers, who have become very conversant with fitting linearly structured models, be they univariate, multivariate or recurring measures, would obviously note the usefulness of GLM procedures (Chatfield & Collins, 1980, p.107). Basic GLM features include sum-of-the-square, estimated-marginal-means, profile plots, as well as custom-hypothesis tests which are optional in 4 approachable measures and are well structured for the evaluation of sums-of-squares (SS). These SS options are quite easy to access. The first type of SS enables the calculation of reduced error SSs through the addition of effects in the model periodically.
The use of multivariate methods in conducting surveying measurements by researchers has proved to be vital in analyzing index constructions as well as in exploring initial-stage data that has specified surveyed subdivisions (Babbie, 1998, p.12; Manly, 1994, p.29). With a good understanding of multivariate methods of sampling, one will certainly appreciate the significance of using such in the determination of index constructions in a very practical angle of consideration. For researchers in the survey, the tool may however not be quite familiar as they are likely to have inadequate knowledge of its usage.
Reference List
Babbie, E. (1998). Survey Research Methods (2nd ed.). Belmont: Wadsworth.
Chatfield, C., & Collins, A.J. (1980). Introduction to Multivariate Analysis. London: Chapman and Hall.
De Vaus, D. A. (1990). Survey in Social Research (2nd ed.). London: Unwin Hyman.
Everitt, B.S., & Dunn, G. (2001). Applied Multivariate Data Analysis. London: Arnold.
Everitt, B.S., Landau, S., & Leese, M. (2001). Cluster Analysis. London: Arnold.
GenStat,. (2002). GenStat for Windows, 6th Edition. Oxford: VSN International Ltd.
Krzanowski, W.J. & Marriott, H.C. (1994a). Multivariate Analysis, Part 1. London: Arnold.
Krzanowski, W.J. & Marriott, H.C. (1994b). Multivariate Analysis, Part 2. London: Arnold.
Manly, B.F. (1994). Multivariate Statistical Methods: A primer, 2nd ed. London: Chapman and Hall.
Oppenheim, A. N. (1992). Questionnaire Design, Interviewing and Attitude Measurement. London & NY: Continuum.
Sahn, D.E. & Stifel, D. (2000). Assets as a measure of household welfare in developing countries. Washington: Washington University.