Research Design and Statistical Critique

This assignment focuses on the critiquing the method and statistical analysis for an existing empirical journal article and should include your peer feedback from this unit's Critique Method and Statistical Analysis: Peer Review discussion. It is an opportunity for you to concisely demonstrate the mastery of the competencies of empirical research or statistical thinking, critiquing, and proposing new research designs.In this assignment, you will use the journal article you worked on in this unit's Critique Method and Statistical Analysis: Peer Review discussion. The methodology of the journal article should be a statistical research design discussed in this course.Complete the following:• Include feedback received from your peers.
• Justify the measurement scheme. Defend that the measures are valid representations of the constructs and their functional roles in the theory.
• Discuss sampling methods and analyze the research design.
• Create and present your own template for critiquing:
o Major research question.
o Research design.
o Variables and measurement.
o Hypothesis.
o Sampling protocol and data collection.
o Exploration and testing of statistical assumptions.
o Selection of statistical test, the results of one omnibus hypothesis, credibility, and reliability.
o Costs, benefits, and limitations of the research.
• Write another paragraph that:
o Suggests a new research question and possible statistical design to advance research on this topic.
o Includes a promising research venue to which you might have access and one where you will not have an ethical conflict of interest.
Save your file as YourLastName10.docx. Submit a maximum of two pages in Microsoft Word, adhering to APA sixth edition style guidelines.
Journal Article Selection
Consider the following while selecting a journal article:
• As most research designs are significantly more complex than those executed in our SPSS labs, you may need to critique only a portion of the article as you have done in Unit 1, which had three different empirical research designs.
• When possible start with the omnibus hypothesis. For example in a multiple regression, the null would be:
o There is no relationship of the independent variables to the dependent variable.
o The F test is the overall statistic. You would report the results of that test first.
o If it is rejected, then the follow-up tests for each of the variables would be conducted. It is highly probable that a more advanced regression analysis was used than that covered in this course. Please report the name of that type of regression. For example, hierarchal regression, logistics regression, et cetera.
o If you select an article using SEM, select only one model tested and report only the root mean square error of approximation.
o If you select an experimental design with more than one dependent variable using MANOVA, report the high-level hypothesis of testing the overall relationship of the variables in that statistical model.
• Regardless of the technique, be sure to include insight into the proper use of probability sampling and examination of each of the assumptions of the statistical model proposed.

Powered by