Research is a process that involves the gathering of relevant information in an attempt to answer some pertinent questions regarding an issue of concern (Salkind, 2010, p.45). Sometimes, the process of research can be quite complex. When this is the case, there arises a need to employ the use of scales or other specific tests by scientists. This is for the simple reason that in so doing, it allows for more accurate and reliable measurement of multiple or interrelated items. Since behavior that is observable may sometimes involve several forces acting together, scales and special tests address how to handle the data in order to maintain or increase precision, validity and reliability of the measurement of data.
Scaling is a common terminology that is used mainly in social sciences to mean the entire process of ordering or arranging various entities based on their character traits. There are different types of scaling and each is most suitable for certain kinds of research plans and not for others. The main types of scaling are comparative and non-comparative. In the comparative kind of scaling, one compares two kinds of items trying to see which is of the two has got more preferences among the audience. For example, a researcher may ask a respondent, between Fanta and Coke, which do you prefer and why? In non-comparative scaling, the researcher looks at the trait of one particular element without comparing it with another. For example, the researcher may ask a respondent, how do you find Fanta?
In this case, the most appropriate kind of scaling will be the comparative one. This is because comparing an element with another, it helps bring out better the traits of whatever it is that the researcher is investigating as opposed to looking at the element solely. Under comparative scaling is Guttman scaling which will also be highly applicable in this case. This kind of scaling is also known as scalogram scaling or analysis (McMurray, 2004, p.23). This kind of scaling aims at enabling the researcher almost predict correctly the kind of responses that the respondents are likely to give based on the response of some of the questions they have already answered. Coming up or developing this kind of scale is very similar to developing any other kind of scale. The first step usually is defining the focus or scope and the next step will involve developing the items which will be a reflection of the concepts.
Reliability and validity are very important concepts when it comes to research (Monsen & Horne, 2010, p.13). A researcher must be able to justify the validity and reliability of the particular scaling method chosen. The validity of the study is fundamental since it concerns the degree of accuracy to which the study correctly assesses, addresses and reflects the concepts being measured. It concerns the success achieved pertaining to whatever the study attempted to measure.
When looking at content validity, the key issue concern is the measurement extent vis a vis the main content field of intention. For instance, the learning skills, interviews and surveys will be analyzed using all relevant mathematical functions since the exclusion of some mathematical functions will reduce the content validity of the study. The data collected will be analyzed using appropriate levels of measurement for validity purposes. Since this is a public health study, it will be necessary to research what constitutes a relevant domain of the prevalence of HIV/AIDS in Africa. This means that the domains involved will be adequately defined before being practically measured to enhance content validity.
Construct validity is about seeking agreement between the specified measuring procedure in the study and a corresponding theoretical principle. For instance, the study will ensure construct validity by defining the causes of the high prevalence of HIV/AIDS in Africa before testing the actual hypotheses. There will be verification of the procedures that will be used in the study with those that are indicated in literature for convergence purposes. To enhance the construct validity of this particular study, the relationship between the procedures employed in the study and the theoretical concepts found in the literature will be specified and verified. There will also be a strict examination of the empirical relationships existing between various concepts that are under investigation in the study. It is also important to ensure that evidence in terms of facts and figures is interpreted especially in relation to making clarification on the construct validity of the particular measure that is being tested or investigated.
The reliability of the study will only be enhanced if the measurements and test scores obtained will prove to be consistent. If the measurements and the test scores will show repeatability, then the study will be termed as reliable. The study will use appropriate means and methods to estimate its reliability. Among the methods that are best applicable in the enhancement of reliability are the test and retest method. In this method, the measurement instrument will be implemented at two different and separate times for various subjects after which the correlation between the measures will be computed. Assuming [there is] no change in the underlying conditions under which the concept is being measured, the procedure will be repeated severally. The results for the test need to be approximately similar in order to enhance the reliability of the study
The internal consistency of the study will be enhanced by grouping together questions that measure the same concept. For instance, the questions intended to address the issue of possible reduction of HIV/AIDS infection will be grouped together. This will be followed by a collection of responses after which it will be important to run a correlation between the two groups each of which has a number of questions. This will help determine the reliability of the level of the instrument being used through evaluation of the consistency of the correlation results. Although the two approaches will ensure the reliability of the study, the only difference is that retest is associated with two administrations while the internal consistency is associated with one administration of the instrument in question (Rubin & Babbie, 2009, p.145).
If a researcher is not able to justify the validity and reliability of the scale in use, then there are ways of determining these two important aspects of scaling. When preparing a quantitative research plan testing is of great importance (Black, 1999, p.21). There are different types of tests that a researcher can use but one must always ensure that the test selected is one that is appropriate. A test in this case refers to an instrument that is standardized which is used to assess the constructs in any kind of research. One kind of test that can be used in this case is a questionnaire. This is a set of questions formulated and designed by the researcher with the aim of extracting crucial information from the respondents. The importance of carrying out tests will ensure that the kind of assessment done is proper. In this case, the most appropriate test will be questionnaires and interviews so as to get to establish the actual information on the ground.
When carrying out research it is hardly possible to use the entire population. This is because it could be that the population being studied is too large and therefore surveying the whole population is costly and may bring about unreliable results.
Therefore researchers usually look for only a small portion of the population which is supposed to have all the demographics of the population thereby ensuring that the portion selected is representative of the entire population. In this case, therefore, the population to be used will only be a small fraction of the entire population which will serve as a representative sample of the entire. This part of the population will be selected using the stratified method of sample selection. A quantitative research plan can be complex but not so much if one understands the intricacies of each and every step. Measurement and instruments for a quantitative research plan need to be understood in order to come up with a proper plan.
References
Black, T. (1999). Doing quantitative research in social sciences: an integrated approach to research designs, measurement and statistics. London: Sage.
McMurray, A. (2004). Research: a commonsense approach. New York: Cengage Learning.
Monsen, E. & Horn L. (2010). Research: Successful approaches. New York: American Dietetic Association.
Rubin, A. & Babbie, R. (2009). Essential research methods for social work. New York: Cengage Learning.
Salkind, N. (2010). Encyclopedia of research design. London: Sage.