RegressionAnalysis andChoosing the Statistical Test to ResearchScenarios
Introduction: SPSS Assignment 7 consists of running several simple, bivariate, linear regression using the College Score Card Data Set as well as determinign the appropriate statistical test based on a research scenario.
Bivariate (Simple) Regression- good explanation of output including a scatterplot Bivariate Linear Regression- good explanation of output Quick Overview of Bivariate Regression- useful to pause and write your notes
Choosing a Statistical Test: These resources include some test we have not covered, just igonore those:
Choosing The Right Statistical Analysis - Resource 1Choosing the Right Statistical Analysis - Resources 2
Task 1: SPSS Regression Analysis (12 points)
Determine if a significant linear model can be created to describe the relationship between the in-state tuition (TUITIONFEE_IN) and the following explanatory variables. NOTE: Run a separate regression analysis for the DV and each IV.
- Family Income (FAMINC)
- % First Generation (PCT_FIRST)
- % PELL (PCTPELL)
- Average Faculty Salary (AVGFACSAL)
- Write a brief introduction. It should begin with a brief background statement that leads to the purpose and introduces the data set followed by the purpose/broad research question(s) guiding the overall analysis.
- Include a brief descriptive write-up of the variables as a whole (not the groups being compared) (e.g. mean, SD, skewness, kurtosis and frequencies, as appropriate based on the variables level of measurement). In other words, paint a picture of the overall data set being analyzed for only the variables being used in the assignment. You may take this information from prior submissions.
- Write a research question and the corresponding null hypothesis being analyzed for each regression analysis. These can be bulleted together followed by the regression results.
- Determine if a statistically significant linear model can describe the relationship between the variables (specifically the assumption that the explanatory variable influences the outcome variable and not vice versa). Be sure to address the test assumptions. Include a scatterplot for each of the analyses.
- What conclusions and recommendations might you draw based on your findings for each regression? Think of this section as a brief presentation of the results, remember readers often skim/skip the analysis part, that tells a story about the results in the context of the research question(s). (Be very careful to not suggest cause and effect type situations). Also include a statement about what you might recommend for future research.
Task 2: Choosing the Statistical Test to Research Scenarios (add to the end of SPSS Assignment) (8points):
For each of the following research scenarios write the null hypothesis and identify the appropriate statistical test. Include a by a brief description of why you selected that statistical test.
- A researcher recorded the calories burned by subjects during a 10-minute treadmill exercise. All participants had television located in front them while exercising. In one group, the television played music videos. In a second group, the television played a news channel. In a third group the television was turned off.
- A study was conducted to investigate the relationship between cigarette smoking and illness. The number of cigarettes smoked daily and the number of days absent from working the last year due to illness were determined for 40 individual employees.
- A statistics faculty member has noticed a trend that students exhibiting higher levels of anxiety perform worse on their test. The faculty member’s goal is to identify, based on anxiety, which students need to be given exercises and support to reduce their anxiety levels so their test performance can improve. To do this the faculty member gives his students an anxiety questionnaire before the final exam, and records the score along with the students final exam score.
- A cognitive psychologist believes that memory for pictures is superior to memory for words. To test this hypothesis, the psychologist performs an experiment in which students from an introductory psychology class are used as subjects. Eight randomly selected students view 30 slides with nouns printed on them and another group of eight randomly selected students view 30 slides with pictures of the same nouns. Each slide contains either one noun or one picture and is viewed for 4 seconds. After viewing the slides, subjects are given a recall test, and the number of correctly remembered items is measured.
- The director of a recreation center on a college campus believes that the more physically active a student is the higher their GPA will be. A random selection of 50 students were selected and were rank ordered by GPA with 1 being the lowest GPA and 50 being the highest. The director then tracked the student’s usage of the recreation center for the week, noting how many days in the 7 day period the student visited the center.
- Employee turnover is a common concern among agencies. Stress is often cited as a key reason for increased employee turnover which has substantial negative consequences on the organization. An HR staff member collected data from employees after 3 month of work about how long they planned to remain with the company (in years) and how stressful they rated the workplace (high, moderate, low stress).
- A behavior therapist is regularly asked by her patients how many sessions will they need to attend to help improve their symptoms. The therapist randomly selects 30 patients who recorded the same amount of symptoms at the start of treatment and records the number of sessions attended and the number of symptoms expressed by each patient.
- One behavioral strategy believed to reduce food intake is to eat slower, which presumably provides more time for physiological fullness signals to initiate. To test this possibility two groups were given the same buffet of foods to consume and the amount of calories each individual consumed. One group was instructed to eat fast and the other instructed to eat slow.