Five Features of Quantitative Research Every Social Science Graduate Student Should Know
[ad_1]
Quantitative research can be described as the investigation of a social problem using a number based approach. It involves a researcher collecting statistics, first hand or by secondary sources such as from the church, police, school, hospital or state agency. Also researchers can use survey instruments or questionnaires to collect qualitative data by asking participants to compose their own answers. In this instance, most of the questions will be open ended and participants will get the opportunity to provide reasons for their choice or supply additional information in support of their opinion. Students should have expert knowledge of assumptions,knowledge of research design, knowledge of methods of integrating measurement and data analysis, knowledge of levels of measurement and knowledge of data interpretation. Moreover, graduate students should harmonize or blend these five knowledge areas creatively and feasibly in their research.
In quantitative research assumptions refer chiefly to the characteristics of the data. Other assumptions can be drawn from social theory upon which the research is being anchored. Before collecting data the student should be clear about the assumptions of the data to be used for answering the research question. It is important that the student identify whether the data are distributed normally or not. This will assist in test selection provided a hypothesis is being tested. Therefore, he or she may select a parametric test like the multiple correlation to ascertain whether any relationships between variables exist. Alternatively, the t-test or analysis of variance may be employed to measure differences in means between two or among more than two groups or samples selected independently. If the data are not normally distributed equivalent non parametric tests such as chi square tests of association and the Mann Whitney test of difference will be selected. Students should be able to explain the assumptions of each test and demonstrate to readers that these assumptions have been met. Depending on the purpose of the study; the good student should possess knowledge of assumptions such as normality, randomness, equality of variance, linearity and independence.
Theories have a number of assumptions about correlation, causality and effects of behavior. Knowledgeable students should be able to point out the relevance of each of these to their study. The student researcher should make it known whether the data enable the determination of correlation or causality. Consequently theory selection, an essential epistemological feature of quantitative research, will require the student to demonstrate a perfect fit between theory and hypotheses. In other words, the student would know that theoretical assumptions should influence the selection of hypotheses to be tested.
It is expected that graduate students should have advanced knowledge of quantitative research designs and their applications. They should be capable of defining research design as a series of steps or procedures logically ordered for data collection and analysis. The graduate must explain that research design is the methods and materials employed in executing the study and is analogous to a plan when skillfully or appropriately implemented produces excellent outcomes. It must be clear to graduate students that the most frequently selected designs are descriptive, survey, correlation based and quasi-experimental. Understanding their differences should be easy. For example; descriptive designs, like the population census; are intended to describe demographic characteristics of the population. They enable researchers to assess the amount of demographic change in a population that took place over a specific period.
On the other hand, graduate students should know that surveys capture the perspectives and perceptions of a cross section of the population at a particular point in time. Surveys are similar to a photographer taking a picture of a person, group or object. In selecting a correlation based study, knowledgeable graduate student researchers should aim to measure the relationship between two variables of interest whereas in quasi experimental studies they should be aware that the aim should be to find evidence of causality. While cohort, retrospective and longitudinal designs can be employed in the social sciences, the diligent graduate will know that by far they are less popular than descriptive, survey or correlation based designs. As stated earlier, the graduate student should comprehend the nexus between theoretical or conceptual framework and design and the importance of illustrating the congruence between them.
One of the most significant pieces or knowledge the quantitative research student should possess is the best way of integrating measurement of variables and processes of data analysis. This is a critical step that has significant impact on the results or findings of social inquiries. For instance, a student should know that interval or ratio measurement is appropriate for performing the Pearson product moment correlation or multiple regression tests. Alternatively, the graduate must know that nominal or categorical measures of dependent and independent variables should lead to the selection of chi square tests which do not have the same amount of power as the Pearson product moment coefficient to prove that the null hypothesis should be rejected or retained. A well trained graduate student will know that in the conduct of an exploratory study a large sample of over five hundred may be selected. If the researcher’s intention is to understand the prevalence or incidence of a problem in a community, the student will know that it will make little sense to mull over levels of measurement if or when no hypotheses are to be tested.
It is necessary that graduate students have exceptional knowledge of the four levels of measurement: nominal, ordinal, interval and ratio. They should know that nominal measures capture categorical data on gender, marital status or ethnicity while ordinal data captures information about rank or rating of events such as athletes’ position in a race or the extent to which patients rate the care they obtained from a health provider, for example, a doctor or nurse. The well informed graduate student would find it easy to devise the best way to measure phenomena such as interest, satisfaction and attitude by ordinal means. A well informed graduate will know that in the social sciences it is not so easy to identify interval measures and that temperature is one of the most frequently cited examples used to explain interval measures. The knowledgeable graduate student will know that many statisticians argue that interval measures, though superior to ordinal, do not indicate the absence of a quantity. A well informed student will know that zero degree Celsius does not mean the absence of temperature.
Additionally, the graduate student should be capable of justifying why ratio measures are at the top of the measurement hierarchy due to the fact that zero means complete absence of the variable under consideration. Examples of ratio measures are money, vehicles, books or houses. The informed graduate should be able to explain another advantage, that is that ratio measures enable counting and comparison, for instance someone with one hundred dollars could claim to have twice as much as another with fifty. Failure to understand levels of measurement can be catastrophic if the data are inappropriate for the nature and purpose of the study. Under this circumstance the student may be impelled to reevaluate his study and make amendments as required. It is extremely painful to the student’s investment in time and energy if a study designed to perform parametric tests collected nominal and ordinal data only.
Finally the graduate student should be competent in data interpretation. An excellent understanding of the rules of hypothesis testing will go a long way towards the production of a successful thesis. Knowledge of issues such as the relationship between probability and alpha values, types of error and data presentation is critical to the analysis and interpretation of data that were analyzed. Graduate students should interpret their results consistently on the principle that the null hypothesis should be retained or not rejected wherever the probability value exceeds the alpha value of point zero five. They should know that it means that any observed relationship or difference between means should be rejected on the ground that it is not statistically significant. It will be beneficial for the graduate student to review the results to ensure that a type 1 error is not made by rejecting a true null hypothesis. On the other hand students interpreting their findings should know that if a false null is not rejected, a type 2 error will be made. Conclusively, graduate students should know the benefit of presenting their results using a variety of charts and graphs as appropriate and that each figure should be titled in accordance with the university’s standard or the style of referencing recommended.
Overall students contemplating undertaking quantitative research must self-reflect on whether they agree with the tenets of positivism. Knowledge of positivism, the belief that the analysis of statistics can produce understandings of the factors correlated and causal to human behavior, is essential. The student should consider that some classical sociologists believed that the causes of human behavior can be found in the goings on in society and not in individual members. For instance, the student should be able to explain that one of the founding fathers of sociology argued that rates of suicide fluctuate according to levels of integration and regulation or control in society. The competent graduate student should know that to attain validity and reliability a quantitative study requires problem definition, literature search and critical writing.
It can be argued that mastery of knowledge of assumptions, knowledge of research design, knowledge of integrating measurement and data analysis, knowledge of levels of measurement and knowledge of data interpretation is invaluable to the successful completion of quantitative research for attaining a master’s degree in a social science discipline like sociology psychology, economics or political sciences.
[ad_2]