Author: Anonymous

Which of the following describes the action of the corrugator supercilii?

After running a linear regression, you obtain the following residuals. Which of the L.I.N.E. assumptions are clearly violated:

FitLife is a mobile fitness app that offers workout programs to its users. The analytics team is modeling Monthly Active Users (in thousands) to understand the impact of their marketing efforts. They want to distinguish between users acquired through paid campaigns and their “organic” user base (users who join through word-of-mouth or app store search). The two explanatory variables are: Social Media Ads (X₁): Monthly spending on social media advertising (in thousands of dollars) Push Notifications (X₂): Number of promotional push notifications sent to users each month There are months when the company does not run any ads or send any push notifications. The regression results are below: Variable Beta Coef. (β) Std. Error p-value Intercept 18.75 3.10 < 0.001 Social Media Ads 1.85 0.40 0.002 Push Notifications 0.95 0.28 0.010 Interpret the intercept.

Which of the following best describes the least squares method?

Model A uses only GPA and SAT Score as predictors. Model B adds Essay Score (rated 0–10 by reviewers) as an additional predictor. Both models are fit on the same 400 applicants. AIC (Akaike Information Criterion) is used to compare model fit. Model Predictors AIC Model A GPA, SAT Score 491.6 Model B GPA, SAT Score, Essay Score 480.1 Based on the AIC values above, which model should the admissions office prefer, and why? 

Variable B S.E. p-value Exp(B) GPA 1.98 0.23 < 0.001 7.23 SAT Score 0.13 0.06 0.031 1.14 Constant −7.77 1.06 < 0.001 0.00 An applicant has GPA = 3.6 and SAT Score = 13. Using the regression output, what is the predicted probability of admission for this applicant?  You may use the following approximations: e^0.350 ≈ 1.42e^0.720 ≈ 2.05e^1.048 ≈ 2.85e^1.560 ≈ 4.76

A university admissions office wants to predict whether an applicant will be admitted (1) or rejected (0). They collect data on 400 past applicants and run a binary logistic regression with two predictors:  GPA (on a 4.0 scale) and  SAT Score (in hundreds, e.g., 13 = 1300). About 44% of applicants in the sample were admitted. Below is the SPSS output (Variables in the Equation). Variable B S.E. p-value Exp(B) GPA 1.98 0.23 < 0.001 7.23 SAT Score 0.13 0.06 0.031 1.14 Constant −7.77 1.06 < 0.001 0.00

Interpret the slope coefficient for log(Number of followers)

You are looking for influencers to promote your product and want to estimate the expected number of likes a post will receive. You regress number of likes (in thousands) on: log(Number of followers), standardized post length, and an indicator equal to 1 if the post is published on a weekend. You only use data from influencers with between 50,000 and 100,000 followers, so log(Number of followers) is approximately between 10.8 and 11.5). The regression results are below: Variable β Coef. Std. Error p-value (β = 0) Intercept 3.102 5.676 0.585 log(Number of followers) 8.050 1.326 < 0.001 Standardized Post Length 2.142 0.253 < 0.001 Weekend 1.742 0.555 0.002  

After running a linear regression, you obtain the following residuals. Which of the L.I.N.E. assumptions are clearly violated: