Need help with an assignment in psychological statistics
Overview: In this assignment, you and your research team will begin exploring inferential statistics—the statistical tests researchers use to test hypotheses.
Directions: Complete all five parts of this worksheet.
PART 1: LOOKING FOR DIFFERENCES BETWEEN GROUPSJuanita has been listening to what other researchers are doing with their project. She’s bothered by the variable sex.
“Why are there so few options for sex? Shouldn’t we reflect the multiple options that exist?”
“The General Social Survey has been around for a long time, it measures sex as biological sex at birth rather than asking about gender,” Duante says. “As researchers, we have options to choose research questions that reflect what we want to learn. Based on that question, we decide what data we use or collect to answer the question.”
“So, if what I want to research requires a close look at gender, I could look for other sources of data?” Juanita asks.
“Correct. For this project, like all archival data research, we use data collected by others and cannot control the questions. We go in knowing that the questions might not ask our ideal question or reflect changes in understanding of certain variables. We can note this in our final research report and discuss the implications. Best practice is to create our hypothesis such that the data we have is perfectly suited to the hypothesis.”
After thinking about the data they do have, Juanita wonders of there is a relationship between sex and the number of hours spent on the internet.
Duante offers to run a t-test, using sex to create two independent groups (as the respondents answered either Male or Female—creating two separate groups).
Directions: Create the hypotheses and run the t-test using the Wwwhr and Happy CSV file.
Scoring Criterion: Perform a t-test on the variables sex and wwwhr.
| What would be the null hypothesis for this question? |
|
| What would be the research hypothesis for this question? |
|
Duante reminds the group that a null hypothesis for a t-test states that the difference in group means is zero and an alternative or research hypothesis states that the difference in group means is different from zero.
With the hypotheses in hand, you pull up the data.
If you haven’t already downloaded it, download Wwwhr and Happy CSV file to your computer. Remember where you saved it.
In JASP, select the three blue bars, select open, find where you saved Wwwhr and Happy CSV file.
Select T-Tests, then Independent Samples T-test
Place Wwwhr in the dependent variable box.
Place sex in the grouping variable box.
Under Additional Statistics, place a check mark next to effect size.
Copy and paste the resulting table below.
Directions: Answer the following three questions.
Scoring Criterion: Interpret the results of a t-test on the variables sex and wwwhr.
| If the ⍺ = 0.01, do you reject the null hypothesis? | ☐Yes ☐No |
| Write out the results using APA’s standards for reporting statistical results, including effect size. |
|
| Explain why it is important to check the effect size. | |
| A t-test compares the means of two groups. Explain how using Sex as a grouping variable in JASP gives you two groups. |
|
If Sex can be used to make two groups, can any of the following variables also be used to create two groups? Are any of these binary (choice of two answers) options? Check the appropriate box.
| Variable | YES (variable can be used to create two groups) | NO (variable cannot be used to create two groups) |
| Raclive: Respondents answered Yes or No as to whether they lived in a neighborhood with other races. |
| |
| News: Respondents were given a version of a Likert scale. |
| |
| Wwwhr: Respondents entered the number of hours they are on the internet. |
| |
| Happy: Respondents were given a version of a Likert scale. |
| |
| Life: Respondents were given a version of a Likert scale. |
| |
| Mntlhlth: Respondents entered the numbers of days of poor mental health. |
| |
| Depress: Respondents answered Yes or No question on whether they had been told they had depression. |
|
Amanda wonders if there is a relationship between the mean number of days with poor mental health and whether or not someone lives in a neighborhood with individuals of a different race.
Directions: Create the hypotheses and run the t-test for independent samples.
Scoring Criterion: Perform a t-test on the variables raclive and mntlhlth.
| What would be the null hypothesis for this question? |
|
| What would be the research hypothesis for this question? |
|
If you haven’t already, download the Raclive and Mntlhlth CSV file (save it where you can find it).
In JASP, select the three blue bars, select open, find where you saved Raclive and Mntlhlth CSV file.
Select T-Tests, then Independent Samples T-test.
Place mntlhlth in the dependent variable box.
Place raclive in the grouping variable box.
Under Additional Statistics, place a check mark next to effect size.
Copy and paste the resulting table below:
Directions: Answer the following three questions.
Scoring Criterion: Communicate statistical information in academic language.
| If the ⍺ = 0.01, do you reject the null hypothesis? | ☐Yes ☐No |
| If the ⍺ = 0.05, do you reject the null hypothesis? | ☐Yes ☐No |
| Write out the results using APA’s standards for reporting statistical results, including effect size. |
|
Duante explains to Amanda that t-tests can be used to answer research questions such as “Do psychology students get better grades in statistics than business learners?” The key, he explains, is that t-tests compare two different groups. If what you want to know is whether two different groups of participants differ in terms of one thing, the t-test might be the answer. Of course, it also depends on what types of data you have. Amanda still looks confused.
Directions: Fill in the table below to help Amanda understand t-tests.
Scoring Criterion: Describe t-tests.
| Types of data for the independent variable (called grouping variable in JASP) | Types of data for the dependent variable | This test is used to determine what type of relationship? | |
| Options: nominal, ordinal, interval, ratio (note: some boxes will have just 1 of these options, other boxes will have multiple for a correct response) | |||
| t-test for independent samples |
|
|
|
Amanda wants to know what you would do if you had ordinal data rather than interval/continuous data. You decide to show her by seeing if there was a difference between the general happiness of people living in a neighborhood with other races and those who do not.
Directions: Create the hypotheses and run the Mann-Whitney.
Scoring Criterion: Perform a Mann-Whitney on the variables raclive and happy.
| What would be the null hypothesis for this question? |
|
| What would be the research hypothesis for this question? |
|
If you haven’t already, download the Raclive and Happy CSV file (save it where you can find it).
In JASP, select the three blue bars, select open, find where you saved Raclive and Happy CSV file.
Select T-Tests, then Independent Samples T-test.
Remove the check next to Student.
Place a check next to Mann-Whitney.
Place happy in the dependent variable box.
Place raclive in the grouping variable box.
Copy and paste the resulting chart below.
Directions: Answer the following three questions.
Scoring Criterion: Interpret the results of a Mann-Whitney test.
| If the ⍺ = 0.01, do you reject the null hypothesis? | ☐Yes ☐No |
| If the ⍺ = 0.001, do you reject the null hypothesis? | ☐Yes ☐No |
| Next, assuming that the alpha level was .05, write out the results using APA’s standards for reporting statistical results, including effect size. |
|
Juanita caught the last part of your demonstration and wants to see it from the beginning. You choose to look at whether there is a difference in how exciting people find life if they live in a neighborhood with individuals of another race or not.
Directions: Create the hypotheses and run the Mann-Whitney.
Scoring Criterion: Perform a Mann-Whitney on the variables raclive and life.
| What would be the null hypothesis for this question? |
|
| What would be the research hypothesis for this question? |
|
If you haven’t already, download the Raclive and Life CSV file (save it where you can find it).
In JASP, select the three blue bars, select open, find where you saved Raclive and Life CSV file.
Select T-Tests, then Independent Samples T-test.
Remove the check next to Student.
Place a check next to Mann-Whitney.
Place life in the dependent variable box.
Place raclive in the grouping variable box.
Copy and paste the resulting chart below.
Directions: Answer the following three questions.
Scoring Criterion: Communicate statistical information about a Mann-Whitney in common language.
| If the ⍺ = 0.01, do you reject the null hypothesis? | ☐Yes ☐No |
| If the ⍺ = 0.001, do you reject the null hypothesis? | ☐Yes ☐No |
| Explain to Amanda what the results mean using common (everyday) language. |
|
Juanita wants to see if there is a difference in how much news is read depending on whether someone has been told they have depression or not.
Directions: Create the hypotheses and run the Mann-Whitney.
Scoring Criterion: Perform a Mann-Whitney on the variables news and depress.
| What would be the null hypothesis for this question? |
|
| What would be the research hypothesis for this question? |
|
If you haven’t already, download the News and Depress CSV file (save it where you can find it).
In JASP, select the three blue bars, select open, find where you saved News and Depress CSV file.
Select T-Tests, then Independent Samples T-test.
Remove the check next to Student.
Place a check next to Mann-Whitney.
Place news in the dependent variable box.
Place depress in the grouping variable box.
Copy and paste the resulting chart below.
Directions: Answer the two questions below and then write out your results.
Scoring Criterion: Communicate Mann-Whitney results in academic language.
| If the ⍺ = 0.01, do you reject the null hypothesis? | ☐Yes ☐No |
| If the ⍺ = 0.05, do you reject the null hypothesis? | ☐Yes ☐No |
| Next, write out the results using APA’s standards for reporting statistical results including effect size. |
|
Amanda is confused about what a Mann-Whitney test is and does.
Directions: Fill in the table below to help Amanda understand Mann-Whitney.
Scoring Criterion: Describe Mann-Whitney.
| Types of data for the independent variable (called grouping variable in JASP) | Types of data for the dependent variable | This test is used to determine what type of relationship? | How is a Mann-Whitney different from a t-test? | |
| Options: Nominal, ordinal, interval, ratio (note: some boxes will have just 1 of these options, other boxes will have multiple for a correct response) | ||||
| Mann-Whitney |
|
|
|
|
Duante is tired of the chit chat and wants to focus on the work at hand. He wants to know about the project you’ll be working on.
Directions: Answer each of the questions in the table below.
Scoring Criterion: Determine whether a t-test or Mann-Whitney can be used for the study.
| What variables are you using for your study? See Part 1 of your Week 3 assignment. | |
| Are any of the variables you are using for your study nominal? If yes, which ones? | |
| If yes, does the nominal variable have two or more response options? | |
| Are any of the variables you are using for your study ordinal? If yes, which ones? | |
| Are any of the variables you are using for your study continuous (interval or ratio)? If yes, which ones? | |
| Putting it together, to use a t-test you need (a) one variable that can be used to create two groups and (b) one variable with an appropriate mean. | |
| So, is there a way of comparing the means of two groups (and therefore being able to use a t-test) for your data? Remember to check for a normal distribution. To check for a normal distribution, in JASP, select Normality under Assumption checks. If the p-value is greater than 0.05, you do have a normal distribution. | ☐Yes ☐No |
| Mann-Whitney offers a test for two groups when the dependent variable is not normally distributed or is ordinal. To check for a normal distribution, in JASP, select Normality under Assumption checks. If the p-value is less than 0.05, you do not have a normal distribution. Does your data lend itself to a Mann-Whitney test? | ☐Yes ☐No |
| Putting it together, t-tests and Mann-Whitney compare two different groups. They work for research questions that want to know the difference between groups. | |
| What is your research question? | |
| Based on your variables and research question, would a t-test be appropriate for your project? | ☐Yes ☐No |
| Based on your variables and research question, would a Mann-Whitney test be appropriate for your project? | ☐Yes ☐No |
Note: In each assignment, we’ll look at the statistical tests we’ve learned about and whether they align with your variables and research question. So—whether you answered Yes or No here, we may learn about a better test to use later on in the course.
PART 5: MEASUREMENTS AND PARTICIPANTSYou gather your team to gather information about your measurements and participants. Duante gives you the websites you need to look at.
Directions: Complete the tables below.
Step 1: Describe the participants and measures for the data you are using.
Duante’s Notes:
The data came from GSS and general information can be found at the About the GSS web page. Specific information about the methodology can be found in Appendix A: Sampling Design & Weighting of the GSS Codebook.
For the participants, include how many participants. For the measures, describe the question(s) and answers possible for the variables you pulled. All of the data is from survey year 2018.
Fill in the following:
| Identify the number of participants (based on your data). | |
| Describe the survey question and possible responses (from GSS). | |
| Briefly describe GSS. |
Step 2: Determine the reliability and validity of measures.
Scoring Criterion: Determine the reliability and validity of measures.
| Discuss the reliability and/or validity of the General Social Survey (GSS). See Hout and Hastings (2016) from the reading list linked in the assignment to get you started as well as Lameck (2013). |
Step 3: Determine the generalizability of specific findings based on parameters of the research design.
Scoring Criterion: Determine the generalizability of specific findings based on parameters of the research design.
| Based on what you discovered about GSS, describe who the results of your study can be generalized to and explain why. Write one paragraph. |
13
