STAT660-01_f18-team-2_project2_data_analysis_by_WH.sas

*******************************************************************************;
**************** 80-character banner for column width reference ***************;
* (set window width to banner width to calibrate line length to 80 characters *;
*******************************************************************************;

*
This file uses the following analytic dataset to address several research
questions regarding the effectness of a direct marketing campaigns of a 
Portuguese banking institution.

Dataset Name: bank_analysis created in external file
STAT660-01_f18-team-2_project2_data_preparation.sas, which is assumed to be in 
the same directory as this file.

See included file for dataset properties
;

* environmental setup;

* set relative file import path to current directory (using standard SAS trick);
X "cd ""%substr(%sysget(SAS_EXECFILEPATH),1,%eval(%length(%sysget(SAS_EXECFILEPATH))-%length(%sysget(SAS_EXECFILENAME))))""";


* load external file that generates analytic dataset bank_analysis;
%include '.\STAT660-01_f18-team-2_project2_data_preparation.sas';


*******************************************************************************;
* Research Question Analysis Starting Point;
*******************************************************************************;

title1
'Research Question: Who were the clients of this bank?'
;

title2
'Rationale: This would help to identify the client of this bank, helping to understand who would be the target clients.'
;

footnote1
'Bank user demographic data such as age,job,marital,education,housing and loan are used to draw a whole picture of bank users.'
;

footnote2
'The average age of the bank users are 40 years old while almost half of them are from 28-38 years. Sixty percent of them are married and most of them do not have a loan.'

;

footnote3
'Based on the results, almost half of the bank users are admin and blue-collar.Besides, there are sixteen percent of them are techinician.'

;

footnote4
'Thiry percent of the users have a university degree while twenty-three percent are in high school degree.'

;

footnote5
'According to the result, half of them have a house while almost another half do not.'

;

*
Note: This is essentially trying to take a look at the users variables age,
job,marital,education,housing and loan in the original bank_subscriber 
and bank_nonsubscriber datasets.

Methodology: Used proc freq to find the frequency of custmoer demographic
data, using the result to draw a picture of the customers of the bank.

Limitations: Since there are limited demographic variables in the dataset,
it would be better to include more attributes.

Follow-up Steps: Add more client demographic data into the data set or 
combine with other data results of bank clients analysis.
;


proc freq 
        data=bank_analysis ORDER=FREQ
    ;
    table 
        job education 
        / maxlevels=5
    Plots=FreqPlot(orient=horizontal scale=percent)
    ; 
run;

proc freq 
        data=bank_analysis ORDER=FREQ
    ;
    table 
        housing loan marital age_range 
        /
    Plots=FreqPlot(orient=horizontal scale=percent)
    ; 
run;



proc means 
        data =bank_analysis 
        mean median maxdec=2
     ;
     var 
        age
     ;
run;

title;
footnote;


*******************************************************************************;
* Research Question Analysis Starting Point;
*******************************************************************************;

title1
'Research Question: Would the age of bank users affect the campaign result?'

;

title2
"Rationale: Based on the analysis before, we noticed that the average age of the bank users are 40 years. This would help to find out what's the relationship between users'age and campaign result which could help to improve the campain performance."

;

footnote1
"A logistic regression model was built to check how the age of users affect their decision of subscription."

;

footnote2
"As user age increase, it is more likely that the campaign would be successful but the probability is relatively small."
;

footnote3
"Based on the result, a small pvalue indicate that user age did affect their final decisions in the campaign even though the affectness is relatively small."

;

*
Note: This compares the columns of age attributes to the outcome of 
subscription colume y in data_subscriber and data_nonsubscriber datasets.

Methodology: A logistic regression model is used to find out the 
relationship. 

Limitations:Since user age is only one of the social and economic context 
attributes.This analysis result may not be enough to undersatand how would 
the social and economic context affect the marketing campaign.

Follow-up Steps:Include more user attributes in the model.
;

proc logistic 
        data = bank_analysis
    ;
    model 
        y (event = 'yes')= age
    ;
run;


title;
footnote;


*******************************************************************************;
* Research Question Analysis Starting Point;
*******************************************************************************;


title1
"Research Question:Is the monthly consumer confidence index and consumer price index related to the campaign result?"
;

title2
"Rationale: This would help understand how did social economic affect the marketing campaign result."
;

footnote1
"A logistic regression was built to check the relationship between the monthly consumer confidence index, consumer price index and the campaign result."

;

footnote2
"The small p values tell us that the relationships are significant in reality."

;

footnote3
"The consumer confidence index has a positive relationship with the campaign result while monthly consumer price index has a negative effect."
;


*

Note:This compares the column ‘y’ from bank_nonsubscriber, bank_subscriber 
and the column ‘consumer confidence’ and 'consumer price' from bank_se data 
set.

Methodology: Use PROC logistic to run the logistic regression to find out 
the exact relationship and whether this is significant in reality.

Limitations: Besides the economics attributs,there are also social attributes 
which may have effections on the campaign result at the same time.

Follow-up Steps: Conduct the analysis with all the possible social/economics 
attributes.
;

proc logistic
        data=bank_analysis
    ;
    model
        y (event ='yes')= cons_price_idx cons_conf_idx
    ;   
run;

title;
footnote;