Week 8 - Signature Assignment: Develop a Statistical Analysis Plan (20 Points)
Week 8 - Signature Assignment: Develop a Statistical Analysis Plan (20 Points)
To complete this Signature Assignment, you will use the statistical knowledge and scientific writing skills acquired in this course. You will be initially asked to develop two types of research questions, or research questions involving relationships between variables and research questions involving the difference between variables.
A research question is a clear inquiry asking just one question and eliciting more than a simple yes or no response. Begin your research question with how, why, or what rather than does, do, or can. Once your research questions have been developed, you will need to create corresponding null and alternative hypotheses for each research question. Remember, the null hypothesis indicates a non-significant relationship or difference will exist while an alternative hypothesis indicates a significant relationship or difference will exist.
Research questions and hypotheses are divided into two categories: relationship questions and descriptive questions.
Relationship research questions include comparison (difference questions) and strength of association (associational):
- Difference research questions are used when comparing scores (on the dependent variable) on two or more groups. These questions attempt to demonstrate that groups are not the same on the dependent variable.
- Association research questions are used when the researcher wants to examine the association or relationship between variables. The approach is commonly used to see how two or more variables covary (vary together). For example, do higher values on one variable correspond to higher values on the other? These questions are also used to how one or more variables can be used to predict another variable.
Descriptive research questions do not use inferential statistics. Descriptive questions are used to summarize or describe the current data. No generalization to a larger population of individuals.
Please review the schematic diagram below:
For example:
Research Question 1:
Q: What is the relationship between family socioeconomic level and student state-mandated test performance?
Ho: A non-significant relationship will exist between family socioeconomic level and student state-mandated test performance.
Ha: A significant relationship will exist between family socioeconomic level and student state-mandated test performance.
Research Question 2:
Q: What is the relationship between family socioeconomic level, gender, and student state-mandated test performance?
Ho: A non-significant relationship will exist between family socioeconomic level, gender, and student state-mandated test performance.
Ha: A significant relationship will exist between family socioeconomic level, gender, and student state-mandated test performance.
Research Question 3:
Q: What is the effect of an intensive language immersion program on spoken English acquisition skills among high school ESL students?
Ho: An intensive language immersion program will result in a non-significant effect on spoken English acquisition skills among high school ESL students.
Ha: An intensive language immersion program will result in a significant effect on spoken English acquisition skills among high school ESL students.
Research Question 4:
Q: What is the effect of an intensive language immersion program on spoken English acquisition skills among high school ESL students by gender?
Ho: An intensive language immersion program will result in a non-significant effect on spoken English acquisition skills among high school ESL students by gender.
Ha: An intensive language immersion program will result in a significant effect on spoken English acquisition skills among high school ESL students by gender.
You will then need to explain the type of statistical analysis to be employed to analyze the relationships and effects between variables. Keep in mind, a statistical relationship analysis involving two variables often involves a Pearson r correlation or a simple regression while a statistical relationship analysis involving three or more variables often involves multiple correlations—or what is more widely known as multiple regression. A statistical analysis testing for effect involving two variables often involves some form of a t-test; on the other hand, a statistical analysis of effect involving three or more variables often involves some form of an ANOVA.
Carefully review the scenario and research questions provided below for you to develop the statistical analyses plans.
Scenario:
You are the new director of institutional research at a small state university, and you have been assigned the task of analyzing information for the dean of the School of Education regarding the performance of their undergraduate students on the often-controversial Graduate Record Exam (GRE). Many educators believe the GRE is a poor evaluator of undergraduate performance as well as a poor predictor of graduate school performance. The dean is considering eliminating the GRE from graduate school admissions requirements.
The dean has already collected data on four variables: 1) gender, 2) grade point average (GPA), 3) GRE score, and 4) graduate degree completion frequency. Your job is to develop a proposed analysis to assist the dean to make an informed decision regarding the future use of the GRE.
You should also discuss the assumptions of each test. No data are required to be presented. This is similar to a question that you will encounter in your Doctoral Comprehensive Exams. You should provide information that shows your understanding of the different types of analyses, as well as possible outcomes of the analyses. In addition, you have to include in your discussion the possible conclusions based on the possible results: rejecting the null and not rejecting the null.
Using this information, develop the following foundational components for a proposed analysis. In your proposal, you will compose four research questions and an analysis of your results and recommendations. For each research question, you need to address:
- Corresponding null and alternative hypotheses.
- Type of statistical analysis to be employed to determine the significance.
- Assumptions of each test.
- Explanations of potential outcomes identifying both non-significant and significant relationships as related to both null and alternative hypotheses—you want to actually calculate results.
- Recommendations based on non-significant and significant findings.
The four types of research questions are:
- A relationship research question involving GPA and GRE scores.
- A relationship research question involving gender, GPA, and GRE scores.
- An effect research question involving gender and GRE scores.
- An effect research question involving gender, GRE score, and degree completion frequency.
Finally, complete your analysis plan with a written of your potential results and recommendations for the dean based on your findings.
-
Rating:
5/
Solution: Week 8 - Signature Assignment: Develop a Statistical Analysis Plan (20 Points)