According to Gaur A. and Gaur S., (2009:31), “reliability and validity are two important
characteristics of any measurement procedure”.
4.9.1 Validity
Validity means that the measuring instrument actually measures the property it is supposed to measure (Gaur A. and Gaur S., 2009:31).
“The objective of assessing validity is to see how accurate is the relationship between the measure and the underlying trait it is trying to measure. The first step in assessing validity is called the face validity test. Face validity establishes whether the measuring device looks like it is measuring the correct characteristics. The face validity test is done by showing the instrument to experts and actual subjects and analyzing their responses” (Gaur A. Gaur S., 2009:31).
Validity is about whether the right concept is measured (Gilbert, 2008:217). Face validity was assessed through pre-testing the questionnaire; and by discussing it with the supervisor.
Internal validity regards the degree to which conclusions about causal relationships can be made. This concept seeks to establish confidence in the ‘truth’ of the findings. From this argument, no doubt the current study findings make sense. The findings fulfill the research objectives that this study set out to investigate. Concerning the concept of external validity, this study provides a detailed portrait of the setting in which the research was conducted to give readers enough information to judge the applicability of the findings to other settings. However, it should be noted that the sample was purposive in nature. There are other two important aspects of validity including content validity and construct validity Gaur & Gaur (2009). Content validity refers to the extent to which a measurement reflects the specific intended domain of content (Gaur A. and Gaur S., 2009:31). To establish content validity, the key concepts are defined in this study. The current study measures job satisfaction of students who were currently working, those who had stopped working, and those who were searching for jobs but had not yet got them. All these groups of students had different human rights issues in as far as Swedish labour market is concerned. To establish content validity, the entire domain of the study was defined and assessed to make sure that the instrument truly represents this domain.
Construct validity is one of the most commonly used techniques in Social Sciences. It looks for expected patterns of relationships among variables. Construct validity thus tries to establish an agreement between the measuring instrument and theoretical concepts (Gaur A. and Gaur S., 2009:33). The researcher established a theoretical relationship and examined the empirical relationships. Empirical findings were then interpreted especially in terms of how they clarified the construct validity.
4.9.2 Reliability
Reliability on the other hand is about whether a measure works in a consistent way (Gilbert, 2008: 217). According Gaur A. Gaur S., (2009:31) “reliability refers to the confidence we can
place on the measuring instrument to give us the same numeric value when the measurement is repeated on the same object.” Reliability is the degree to which one may expect to find the
same result if a measurement is repeated. It includes two concepts: stability and consistency. Stability is usually measured by administering the same instrument twice to the same respondents, the time interval being chosen so as to minimize the effects of memory while avoiding the likelihood that real change may have taken place (Gilbert, 2008: 218). Consistency is generally considered more significant. One way to ideally measure reliability is by the test-retest method. It is done by measuring the same object twice and correlating the results. If the measurement generates the same answer in repeated attempts, it is reliable. The researcher tried as much as possible to describe how the analysis was done in methodology chapter. This implies that after running the tests, same results may be derived at. However, establishing reliability through test-retest is practically very difficult.
4.9.3 Generalization
According to Shuttleworth M., (2008), generalization is an essential component of the wider scientific process. In an ideal world, to test a hypothesis, you would sample an entire population. Generalization is applied by researchers in an academic setting. It can be defined as the extension of research findings and conclusions from a study conducted on a sample population to the population at large. This is what Sturgis P., (2008) calls statistical inference
from the sample to the population. The underlying motivation of sampling is to make statistical inferences from the sample to the population. In reality, it is not possible to sample the whole population, due to budget, time and feasibility (Shuttleworth M., 2008). “Populations are often extremely large; it is usually impossible- for cost and practical reasons, to make measurements on every element in the population. For this reason, we draw a sample and generalize from the properties of the sample to the broader population” Sturgis P., (2008:167). In context of this study, the concept of generalization is very relevant. It was impossible to cover all the students at Gothenburg University therefore the sample was selected. Seven departments within the faculty of Social Sciences were purposively selected that represented other students within the same university; this was because of limited resources available to the researcher. Although, the response rate was low, (20 per cent), all students (515 in total) within seven selected departments were given equal chances to participate in this study. However, the findings from this sample cannot be generalized to the entire population since the sample was not randomly selected and given low response rate. The findings only give us a clear picture about the sample.