Homework 5

Ernesto Gomez

October 9, 2017

---

This assignment performs a simple regression to understand the possible influence of interest in celebrity lifestyle, fear of ageing, and gender on whether one considers themselves to be big spenders on their appearance.

A group-wise summary and visuals will be included to verify regression results.

---

Data Management

Variables

• spendlook - This variable will be our dependent variable. The variable measures whether a young respondent claimed to be someone who spent a lot of money on their appearance on a 5-point Likert scale and treated as continuous (1 Strongly Disagree -> 5 Strongly Agree) - “I spend a lot of money on my appearance”
• celebrity - This variable measures whether a young respondent is or is not interested in celebrities lifestyles on a 5-point Likert scale (1 Not Interested ->5 Very Interested) - “Celebrity lifestyle”
• fearageing - This variable measures whether a young respondent fears or does not fear ageing/growing older on a 5-point Likert scale (1 Not Afraid at all -> 5 Very Afraid of): “Ageing”
• gender - This variable measures whether a respondent is male or female, 2-point scale, categorical

library(tidyverse)
library(dplyr)
library(sjmisc)
library(radiant.data)
library(pander)
library(Zelig)
library(texreg)
library(visreg)
yp2 <- data.frame(read_csv("/Users/ernesto/Documents/Advanced Analytics/Data/youngpeople.csv")) %>%
 rename("fearageing" = Ageing,
    "spendlook" = Spending.on.looks,
    "gender" = Gender,
     "celebrity" = Celebrities) %>%
  select(celebrity, gender, fearageing,spendlook) %>%
  filter(!is.na(spendlook),
         !is.na(fearageing),
         !is.na(celebrity),
         !is.na(gender), 
         gender %in% c("male", "female")) %>%
           mutate(gender=as.factor(gender))
head(yp2)

---

Regressions

Model 1

Results indicate an a statistically significant relationship between interest in celebrity lifestyles and fear of ageing on spending on appearance.

• For every one unit increase in agreement with the variable celebrity there is an increase by .25 in estimated agreement with the statement “I spend a lot on my appearance” in spendlook
• Similarly, for every one unit increase in agreement with the variable fearageing there is an increase by .13 in estimated agreement with the statement “I spend a lot on my appearance” in spendlook.
• Gender showed no significance, with men showing a decrease in agreement with spending a lot on appearance.

lm1 <- lm(data = yp2, spendlook ~ celebrity + fearageing + gender)
summary(lm1)

Call:
lm(formula = spendlook ~ celebrity + fearageing + gender, data = yp2)

Residuals:
     Min       1Q   Median       3Q      Max 
-2.86127 -0.77787  0.00791  0.89206  2.48048 

Coefficients:
            Estimate Std. Error t value             Pr(>|t|)    
(Intercept)  2.22872    0.11015  20.234 < 0.0000000000000002 ***
celebrity    0.24667    0.02970   8.306 0.000000000000000322 ***
fearageing   0.12917    0.02669   4.840 0.000001504640455971 ***
gendermale  -0.08505    0.07647  -1.112                0.266    
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 1.141 on 994 degrees of freedom
Multiple R-squared:  0.1098,    Adjusted R-squared:  0.1071 
F-statistic: 40.87 on 3 and 994 DF,  p-value: < 0.00000000000000022

---

Model 2

Interaction shows no significance; men with interest in celebrity lifestyles show an increase in likelihood to be big apperance spenders by .01.

lm2 <- lm(spendlook ~ celebrity*gender + fearageing, data=yp2)
summary(lm2)

Call:
lm(formula = spendlook ~ celebrity * gender + fearageing, data = yp2)

Residuals:
     Min       1Q   Median       3Q      Max 
-2.85677 -0.77222  0.00376  0.89935  2.48553 

Coefficients:
                      Estimate Std. Error t value             Pr(>|t|)    
(Intercept)           2.236845   0.125829  17.777 < 0.0000000000000002 ***
celebrity             0.243884   0.036283   6.722      0.0000000000302 ***
gendermale           -0.103482   0.157608  -0.657                0.512    
fearageing            0.128877   0.026795   4.810      0.0000017447688 ***
celebrity:gendermale  0.008344   0.062369   0.134                0.894    
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Residual standard error: 1.142 on 993 degrees of freedom
Multiple R-squared:  0.1098,    Adjusted R-squared:  0.1062 
F-statistic: 30.63 on 4 and 993 DF,  p-value: < 0.00000000000000022

htmlreg(list(lm1, lm2))

Statistical models

	Model 1	Model 2
(Intercept)	2.23***	2.24***
	(0.11)	(0.13)
celebrity	0.25***	0.24***
	(0.03)	(0.04)
fearageing	0.13***	0.13***
	(0.03)	(0.03)
gendermale	-0.09	-0.10
	(0.08)	(0.16)
celebrity:gendermale		0.01
		(0.06)
R2	0.11	0.11
Adj. R2	0.11	0.11
Num. obs.	998	998
RMSE	1.14	1.14
p < 0.001, p < 0.01, p < 0.05

---

Group-wise Summaries

yp2celeb <- yp2 %>%
select (celebrity, fearageing, gender, spendlook) %>%
  group_by(celebrity) %>%
  summarise(mean = mean(spendlook))
print(yp2celeb)

yp2ageing <- yp2 %>%
select (celebrity, fearageing, gender, spendlook) %>%
  group_by(fearageing) %>%
  summarise(mean = mean(spendlook))
print(yp2ageing)

yp2interaction <- yp2 %>%
select (celebrity, fearageing, gender, spendlook) %>%
  group_by(gender,celebrity) %>%
  summarise(mean = mean(spendlook))
print(yp2interaction)

Visuals

visreg(lm2, scale="response")

Conditions used in construction of plot
gender: female
fearageing: 2
Conditions used in construction of plot
celebrity: 2
fearageing: 2

Conditions used in construction of plot
celebrity: 2
gender: female