Research Designs - Assignment Example

Research Designs Research studies provide valuable data that helps in understanding phenomenon and devising applications and treatments. But a badly executed research study can provide contaminated data that can lead to misleading conclusions. Thus, it is necessary that a research study be designed such that it collects the most accurate data and reduces the chances of error to the minimum. A good research design has four component parts, each of which helps in enhancing the quality of the study and reducing errors (Frankfort-Nachmias & Nachmias, 2008). 1) The study allows us to compare conditions such that effects of one variable on another can be verified at various levels.2) It allows for the manipulation of the Independent Variable so that the researcher is able to study as many effects of the said variable as possible.3) It allows for comparison with a control condition where the Independent variable is not introduced.

This helps us to verify that the effects seen are actually due to the said Independent variable and not other factors.4) It allows us to generalize the findings of the research to other situations.A true experimental design has all these components; and thus offers the most reliable data and analysis (Kerlinger, 1986). An example of a true experimental study would be when a new medicine is being tested, and different dosages are given to different groups to test it’s efficacy. A control group in not given the medicine, but a placebo instead.

The dosage levels are varied as per to opinion of the professional researcher; and the sample chosen is such that it represents the larger population from which it comes. On the other hand, a cross-sectional design and a quasi-experimental design would not allow for manipulation of the independent variable or for the presence of a control group given the nature of the Independent variables under study (Frankfort-Nachmias & Nachmias, 2008). An example of a cross-sectional study would be a study that tried to verify the type of music preferred across different age groups.

Although the experimenter can choose the range of each group; there is no ‘control group’ for this study. On the other hand, a quasi-expeimental design would be one where the efficacy of different treatment options for an illness are compared (Frankfort-Nachmias & Nachmias, 2008). The researcher cannot assign treatment options and has to collect data from participants who choose the said options. There can be a control group only if there are people who choose to abstain from treatment. The least strong type of study is one that uses a pre-experimental design (Frankfort-Nachmias & Nachmias, 2008), which only allows for comparison or observation.

A typical pre-experimental design is when a group is tested for knowledge before and after a training procedure and then the two scores are compared for difference.The strength of the research design affects the validity of the study. Validity of research findings has two aspects – internal validity and external validity (Kerlinger, 1986). Internal validity is the extent to which a study is able to demonstrate the true nature of the relationship shared by the variables studied (Kerlinger, 1986).

this kind of validity can be easily affected by the History of the participants – their past experiences which affect their expectations – and their Maturation – or learning that occurs naturally – during the study (Frankfort-Nachmias & Nachmias, 2008). External validity on the other hand, is the concern about the extent to which the data is representative of the population and thus, it’s generalisablity (Frankfort-Nachmias & Nachmias, 2008). When internal validity is threatened, the values of the findings comes under question, while a threat to external validity affects the chances of understanding the population on the basis of the sample (Kerlinger, 1986).

ReferencesFrankfort-Nachmias, C. & Nachmias D. (2008). Research Methods in the Social Sciences. (7th ed). New York: W H Freeman & Co.Kerlinger, F. (1986). Foundations of Behavioral Research (3rd ed.). New York: Holt, Rinehart & Winston.