Reference no: EM132252573

Part I: Problems

Problem 1 - On CourseSite you will find a dataset called "Midterm_problem1." The data comes from a study by the University of Washington looking at the efficacy of an intervention aimed at motivating binge drinking college students to drink less. Participants were randomly assigned to one of two conditions (Intervention, control = 0, treatment =1). The treatment condition received a normative feedback intervention about college drinking. We are going to look at one outcome of the study, drinking and driving (variable name DUI: 0 = no driving under influence, 1 = reported driving under influence). Three months after the intervention, students completed a survey where they reported whether they had driven after drinking 6 or more drinks in a row (DUIvariable), the number of drinks they consume on a typical week in the past 3 months (DRINKS), ability to refuse drinks (DRSE, a higher number indicates that the student felt more comfortable refusing drinks), perceived descriptive norms (Norms, a higher number indicates that they believe the typical students drinks more drinks per week), and perceived attitude of others about drinking (Attitude, higher scores indicate that they think other students will judge them negatively for drinking and driving). We want to examine (1) whether the intervention predicts whether or not someone reports drinking and driving at the 3 month follow-up (DUIvariable), and (2) whether the effects of the intervention on drinking and driving are mediated by how much students drank at follow-up (DRINKS), their ability to refuse drinks (DRSE), perceived norms (Norms), and perceived attitudes of others (Attitude).

A) Write out the regression equation for the model predicting DUI at 3 months from the all independent variables mentioned above.

B) Before conducting the analysis, perform any data checks you think are appropriate (describe any decisions you make based on each data check you perform).

C) Run the analysis and include a copy of your output. Write up your results in APA format (e.g., for an APA journal). Make sure you interpret your results. Make sure to include the following things in your write up: a table with your results, measure(s) of model fit for the regression predicting the main outcome (e.g., DUI), an interpretation of regression coefficients relevant to your research questions, and a description of how you analyzed the data (similar to the format you would see in an APA journal).

Problem 2 - On coursesite you will find a file called "Midterm_problem2" This file contains data from a study examining factors that contribute towards sleep problems. The dependent variable is a composite measure of frequency and severity of sleep problems (Sleep, higher scores indicate more frequent and severe sleep problems). The researchers were interested in whether anxiety (Anxiety, higher scores reflect higher levels of anxiety) and caffeine use (Drink, higher scores indicate more caffeine use during the day) contribute toward problems sleeping after controlling for participant age (Age- this should be included as a covariate). Further, the researchers wanted to know (1) whether using electronic devices (e.g., phone, TVs, iPad, etc.) in bed at night (Electron, 0 = no use, 1 = use) interacted with anxiety to predict sleep problems and (2) whether electronic device use interacted with caffeine use to predict sleep problems. [Note - The only interactions you should include in your model are the Electronic × Anxiety and the Electronic × Caffeine [Drink] interactions]

A) Select an appropriate regression model and test the relevant assumptions. Describe what procedure you use and any decisions you make based on each data check you perform.

B) Run the appropriate analysis and include a copy of relevant output. Write up your results in APA format (e.g., for an APA journal). Make sure you interpret your results. Include appropriate figure(s) and a description of how you analyzed the data (again similar to the format you would see in an APA journal).

Part II: True or False Questions

Indicate whether the following statements are true or false. If you believe the statement is false, explain why in one or two sentences. If you indicate that a statement is false but do not explain why you will not receive credit for that question.

1) In logistic regression, we predict the probability of the dependent variable in order for there to be a linear relationship between the predicted outcome and the predictor(s).

2) In a mediation model looking at whether M mediates the relationship between X and Y, Sam finds a significant positive indirect effect. Sam should conclude that X is related to Y because X is associated with an increase in M, and M is associated with an increase in Y.

3) When using regression to examine moderation, mean centering your variables prior to constructing the interaction terms reduces multicollinearity.

4) Dustin mean centers his independent variable X, and then includes X and its quadratic term (X2) in a regression predicting Y. The regression coefficient for X is b = 0.32 (p < .05) and for X2is b = 0.09 (p < .05). Dustin should conclude that the relationship between X and Y has a concave upward curve.

5) Examining conditional indirect effects allows us to assess whether the effect of X on Y through M depends on some fourth variable.

6) Lucy tells you that the odds of a child having a language delay is 0.09. This means that the likelihood of a child having a developmental delay is 9%.

Attachment:- Assignment Files.rar