Instructions
Remember, 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 ‘what’ or ‘to what extent does’ rather than does, do, or can. Generally quantitative research questions also do NOT start with ‘how’ or ‘why’ because there is not a statistical analysis that can answer all of the reasons why or how something occurs (this is more of a qualitative inquiry). The only exception is ‘how much’ which denotes a numerical answer. Once your research questions have been developed, you will need to create corresponding null and alternative (also known as a research hypothesis) 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 three categories (descriiptive, correlational, differences) that answer three types of claims (frequency, association, or causal).
Descriiptive research questions do not use inferential statistics. These include frequencies, means, percentages, variance, etc. The frequency claims made that are based off of this data can only be used to summarize or describe the data (i.e., the current sample) and cannot generalize to a larger population of individuals.
Correlational research questions are used when the researcher wants to examine the association (i.e., relationship) between variables. The approach is commonly used to see how two or more variables covary (vary together). In other words, do higher values on one variable correlate to higher values on the other? For example, to what extent does studying correlate with grades? If a regression analysis is added, these questions can also be used to determine whether one or more variables can be used to predict another variable. Only association claims can be made from research questions that have correlational research questions. Correlation does not equal causation. We can determine if one variable relates to another variable, but not how they relate and certainly not that one variable causes the other. (Note. We can never use causal verbs when describing correlational research.)
Causal claims can only be claimed after conducting experiments or quasi-experiments. There are many inferential statistics that can be conducted for experiments or quasi-experiments depending on the research question and how the experiment is designed. Difference research questions are used when comparing scores (on the dependent variable) between two or more groups. These questions attempt to demonstrate that groups are not the same on the dependent variable. Differences within groups can be determined through pre- and post-tests analyzed with a paired-samples t-test to analyze whether a significant difference exists before and after a treatment or intervention.
Please review the schematic diagram below:
Figure 11
Schematic Diagram
For example:
Research Question 1 (Correlational):
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 (Correlational with a regression analysis):
Q: To what extent does family socioeconomic level and gender predict student state-mandated test performance?
Ho: Family socioeconomic level and gender does not predict student state-mandated test performance.
Ha: Family socioeconomic level and gender do predict student state-mandated test performance.
Research Question 3 (Causal):
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 does not have a significant effect on spoken English acquisition skills among high school ESL students.
Ha: An intensive language immersion program does have a significant effect on spoken English acquisition skills among high school ESL students.
Research Question 4 (Difference):
Q: Is there a significant difference between students in an intensive language immersion program and students not in an intensive language immersion program on spoken English acquisition skills among high school ESL students?
Ho: There is not a significant difference in spoken English acquisition skills among high school ESL students in an intensive language immersion program and those not in an intensive language immersion program.
Ha: There is a significant difference in spoken English acquisition skills among high school ESL students in an intensive language immersion program and those not in an intensive language immersion program.
Research Question 5 and 6 (Descriiptive):
Q: What is the average test score on the English acquisition test among high school ESL students?
Q: What is the frequency of high school ESL students that are enrolled in an intensive language immersion program?
After you develop the research question and hypotheses, you will need to state the type of statistical analysis to be conducted to answer the research question. Keep in mind, a statistical relationship analysis involving two variables often involves a Pearson r correlation analysis or a simple linear regression (if using ‘predict’ in the RQ), while a statistical analysis involving three or more independent variables often involves multiple correlations—or what is known as multiple regression. A statistical analysis testing for effect or differences typically includes one of the types of t-tests; on the other hand, a statistical analysis of effect or differences 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 I:
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 in making 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. 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 must include in your discussion the possible conclusions based on the possible results: rejecting the null and not rejecting the null, and potential claims you can make based off of these conclusions.
Using this information, develop the following foundational components for a proposed analysis. In your proposal, you will compose three 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.
Recommendations and potential claims based on non-significant or significant findings.
The three research questions should involve:
A relationship research question involving GPA and GRE scores.
A relationship research question involving gender, GPA, degree completion frequency, and GRE scores.
A descriiptive research question involving gender, GPA, degree completion frequency, or GRE scores (no null or alternative hypothesis is necessary for this RQ).
Finally, complete your analysis plan with a written discussion of your potential results and recommendations for the dean based on your findings. Remember, there are additional resources available in the Supplemental Resources under Course Resources from the course home page.
Scenario II:
You are the principal of a middle school and want to know if a new social emotional learning (SEL) curriculum your teachers are wanting to implement is worth the extra cost and resources.
The principal decides to conduct a pilot test for one year. Half of the classes will receive the new emotional learning curriculum and the other half will not receive the SEL curriculum. Data will be collected on four variables: 1) grades in the class, 2) one behavioral measure, 3) stress questionnaire, and 4) SEL skills. Your job is to develop a proposed analysis to assist in making an informed decision regarding the future implementation of the SEL curriculum.
You should also discuss the assumptions of each test. No data are required to be presented. 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 must include in your discussion the possible conclusions based on the possible results: rejecting the null and not rejecting the null, and potential claims you can make based off of these conclusions.
Using this information, develop the following foundational components for a proposed analysis. In your proposal, you will compose three 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.
Recommendations and potential claims based on non-significant or significant findings.
The three types of research questions are:
A between groups differences research question involving students that receive SEL and do not receive SEL on at least one of the outcome variables (DV).
A within groups differences research question involving at least one outcome variable.
A descriiptive research question involving at least one of the variables (no null or alternative hypothesis is necessary for this RQ).
Finally, complete your analysis plan with a written discussion of your potential results and recommendations for the dean based on your findings. Remember, there are additional resources available in the Supplemental Resources under Course Resources from the course home page.
Length: 5-7 pages, not including references and title page. All questions in the assignment need to be answered. In addition to the answers to the assignment questions, also submit the output (.spv) file. (Note. SPSS automatically generates the .spv file as you work in SPSS. When you close your SPSS main window, SPSS will ask you if you want to save the output file. Click ‘yes’, then save to your computer and upload with your assignment.)