Moderation of adolescents do hobbies within AIDS knowledge and current residence stability

Introduction

As said in the link,

The National Longitudinal Study of Adolescent Health (AddHealth) is a representative school-based survey of adolescents in grades 7-12 in the United States. The Wave 1 survey focuses on factors that may influence adolescents’ health and risk behaviors, including personal traits, families, friendships, romantic relationships, peer groups, schools, neighborhoods, and communities. 

source: Data Analysis Tools par Université Wesleyenne 

NB. This study is under Coursera training (Outils d'analyse des données) using python

Data

• H1GI3 (General Introductory)How old was the adolescent when he moved to his current residence?
  It's a quantitative value =>response (1, 2, 3 ... years)

• H1GH45 (General Health)

  How many people do you know who have AIDS?
  

  it's a quantitive value => response: range 0 to 98 people

• H1DA2 (Daily Activities)how many times did adolescents do hobbies? It's a categorical value => response ( not at all, 1 or 2 times, more, refused, don’t know...)

Objective

We study the moderation of did adolescents do hobbies about the current residence period and AIDs existence in neighborhood knowledge.

• N.B As explanatory and response used data are both quantitive, we would like to use Pearson correlation (r correlation analysis).

Python code

# -*- coding: utf-8 -*-"""
Created on Mon Apr 20 00:30:09 2020
@author: ABID Aymen
"""
# CORRELATION
import pandas
import numpy
import scipy.stats
import seaborn
import matplotlib.pyplot as plt
mydata = pandas.read_csv('addhealth_pds.csv', low_memory=False)
mydata['H1GI3'] = mydata['H1GI3'].apply(pandas.to_numeric, errors='coerce')
mydata['H1GH45'] = mydata['H1GH45'].apply(pandas.to_numeric, errors='coerce')
mydata['H1DA2'] = mydata['H1DA2'].apply(pandas.to_numeric, errors='coerce')
#data cleaning from no input values
mydata['H1DA2']=mydata['H1DA2'].replace(' ', numpy.nan)
data_clean=mydata.dropna()
#%%
#pearson information (r,p)
print ('Association between residence stability and AIDS people knowledge  (r,p)')
[r,p]=scipy.stats.pearsonr(data_clean['H1GI3'], data_clean['H1GH45'])
print([r,p])
#moderator grouping
def incomegrp (row):
   if row['H1DA2'] <= 3:
      return 1 
   elif row['H1DA2'] > 6:
      return 2
data_clean['incomegrp'] = data_clean.apply (lambda row: incomegrp (row),axis=1)
chk1 = data_clean['incomegrp'].value_counts(sort=False, dropna=False)
print(chk1)
sub1=data_clean[(data_clean['incomegrp']== 1)]
sub2=data_clean[(data_clean['incomegrp']== 2)]
#pearson information (r,p) per subgroup (sub1..)
print ('Association between uresidence stability and AIDS people knowledge for exact answer')
print (scipy.stats.pearsonr(sub1['H1GI3'], sub1['H1GH45']))
print ('       ')
print ('Association between uresidence stability and AIDS people knowledge for not exact answer')
print (scipy.stats.pearsonr(sub2['H1GI3'], sub2['H1GH45']))
#%%
#graphics
scat1 = seaborn.regplot(x="H1GI3", y="H1GH45", data=sub1)
plt.xlabel('Residence Rate')
plt.ylabel('AIDS people knowledge Rate')
plt.title('Scatterplot for the Association Between Urban Rate and Internet Use Rate for exact answer countries')
print (scat1)
#%%
scat2 = seaborn.regplot(x="H1GI3", y="H1GH45", fit_reg=False, data=sub2)
plt.xlabel('Residence Rate')
plt.ylabel('AIDS people knowledge Rate')
plt.title('Scatterplot for the Association Between Urban Rate and Internet Use Rate for not exact answer')
print (scat2)

Results

Association between residence stability and AIDS people knowledge  (r,p)

[0.15162579814828608, 9.506977908603123e-35]

Association between uresidence stability and AIDS people knowledge for exact answer

(0.034754024175195075, 0.005088079230752868)       

Association between uresidence stability and AIDS people knowledge for not exact answer

(1.0, 0.0)

So,

For the exact answer income group the correlation between 

residence stability and AIDS people knowledge are 0.15 and the P-value is so significant. 

 

For the not exact answer, the association between 

 

residence stability and AIDS people knowledge is 0.034 less

 

 significant P-value

 

.

 

When we map these findings onto the associated scatter plots for 

 

each income group, we are better able to visualize the significant and 

 

non-significant relationships. 

 

Estimating a line of best fit within each scatter plot 

 

shows the positive association between 

 

residence stability and AIDS people knowledge use rate among the exact responses more than not exact (orange points).