Research Questions, Essay Example

What are five elements of a good quantitative design? Discuss design threats to internal validity, external validity, and reliability in a quantitative design? As a researcher, how could you control for these design threats?

In quantitative research, my aim is to determine the relationship between an independent and a dependent variable that is evident in the population. Quantitative research designs can be classified as either descriptive or experimental. In the first case, the subjects are usually measured once, whereas the second case would involve subjects being measured prior to and after a treatment. A descriptive study is based on establishing associations between variables, whereas an experiment establishes causality.

In forming a quantitative design, one requires the five elements that make up quantitative design which are experiment, case-control study,cross-sectional study,quasi-experiment and longitudinal study. (Sun, 2008)

In terms of internal validity, once it has been determined that there is a relation between two variables (A & B), the succeeding issue to be established is based on causality. Does A avertedly cause B? In the case that the study cannot prove internal validity, one cannot establish the cause and effect statements in accordance to the research; the study would be classified as descriptive but not causal. (Cohen, 1988)

Non-random assignment of subject distribution to groups may cause a threat to internal validity. For instance, if a study contains a pretest, along with an experimental treatment, and a follow-up post-test, the element of history is a threat to internal validity (Sun, 2008). If a variable is identified between two age groups, the difference may be down to age or it may be caused by some other variable.

Non-age related aspects, which are inclusive ofgenerational differences  or history related cohort. If a difference is found between the pre-test and post-test, it may be as a result of the experimental treatment, however the cause may also be downto any other event that individuals experienced in the course of the two times of testing.

External validity, on the other hand, addresses the issue of being able to simplify the final data of your study for future cases that will apply the same factors of times, places, and persons. For example, if an individual is to carry out a study that analyzes heart disease in men, will these results be applicable to adolescence or women? Sampling is a constant thorn in developmental research (Cohen, 1988).  Through the years, old age has been perceived to be a period that results in numerous cases of health problems, isolation, depression and severe cognitive deterioration as the research individuals consisted of elderly who resided in nursing homes! Based on this fact, one would be required to make inquires, and in dong so, ask the following questions so as to assess if a threat to the external validity is a possible influence: “Would my conclusion have these identical results with a different sample?”, “Would the results be identical, if I carried out my study in a varied setting?”, and “Would the results remain the same if I had carried out this study in the past or even if I did the same study in the future?” If the result is “yes” to each and every question, then the general validity of my study will be adversely affected.

Identify an appropriate quantitative design, and discuss the strengths and weaknesses of the design.

The true experiment is considered to be the only methodology that can be considered to be a measure of the cause and effect relationship.  Below are examples in which “R” means.

In a random assignment, O will indicate observation, while X means treatment with each line representing a group.

What type of sampling would you use? Explain why you would use this type of sampling.

The sampling I would opt for is the Probability sampling (Representative samples), either random or and stratified.

Random sample implies that each individual that is part of a specified population of interest has an equal probability of selection. It has a very strict meaning. One cannot simply collect responses on the street and have a random sample.

Stratified sample, on the other hand, is a mini-replica of the population. Prior to sampling, the population is categorizedbased on the different characteristics of importance in accordance to the research. For instance, this can be by education, gender, social class, age, etc. Stratified samples are more ideal in comparison to random samples, but they require a large amount of advanced knowledge of the population’s characteristics, thus they are more difficult to create. (Robert, 1998)

Specify at least three inclusion criteria and three exclusion criteria. Explain how a researcher determines these criteria.

Three inclusion criteria are age, sex and race. The three exclusion criteria are any patient whose ailments or disease affects their body temperature. Pregnant women, participating as a subject in any other medical psychological and or biomedical research project; in the case of an individual previously involved as a subject, they must be given a reasonable duration of time so as to rid themselves of the investigational agent.The type of research they are doing, objective of research, the hypothesis of the study and purpose of the research are also very important determinants for a researcher to consider.(Glaser, 1967)

Using G power, determine sample size? What role does the power analysis play?

Large Effect Size: Sample size is based on an alpha of 0.05, with a power of 0.80, a large effect size (d = 0.8), and one tail. There will be an equal distribution of participants in each group. Sample size is 42

Medium Effect Size: Sample size is based on an alpha of 0.05, and is based on a power of 0.80, a medium effect size (d = 0.5), and one tail. There will be an equal distribution of participants into each group. Sample size is 106.

Small Effect Size: Using an alpha of 0.05, a power of 0.80, a small effect size (d = 0.2), and one tail. There will be an equal allocation of participants into each group. Sample size is 650.

References

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). New Jersey: Lawrence Erlbaum.

Glaser, B. & Strauss, A. (1967). The discovery of grounded theory: Strategies for qualitative research. Chicago: Aldine.

Elliot Robert (1998). “Editor’s Introduction: A Guide to the Empirically Supported Treatments Controversy”. Psychotherapy Research

Ron Sun, (2008). The Cambridge Handbook of Computational Psychology. Cambridge University Press, New York. 2008.