Comparison Of Government Assistance And Borrowing Habits By Education Level

1.Compare the percentage of government assistance by each level of education.
2.Compare the percentage of Borrowed in past 12 months from a financial institution and the Borrowed in past 12 months: for education or school fees.

The data has been collected from World Bank as part of global financial inclusion study. The objective of the study is to find the borrowing and the saving hobbits of Australians based on the education level. The results of the analysis might be the participants borrowed and saving hobbits different for the all education levels.

To meet the business objective and provide suggestions of the problem that “the borrowing and the saving hobbits of Australians based on the education level”, the researcher will check the following points to meet the data validity:

1. Whether the data is primary or secondary and limitations of the data.
2. Level of measurements and the analysis collection process.

To provide suggestion for business objective, the researcher will do the following analysis:

1. The SAS software will be used for analysis to meet the accuracy of the results and to read the data into SAS using data step.
2. The analysis of the education level of participants, whether participants are primary or less, secondary and tertiary or above.
3. The One-way ANOVA will be used to check whether there is a difference in the monthly income for new variable Education at 2% level of significance.
4. Compare the percentage of government assistance by each level of education.
5. Compare the percentage of Borrowed in past 12 months from a financial institution and the Borrowed in past 12 months: for education or school fees.  

To import the provided data in to SAS use “Libname” statement and create a permanent library as name “A3”. The SAS codes are given below:

libname A3 "E:WORKNTOCT813313SAS"; run;

 After creating the “A3” library use the following codes to import the data files named as “Borrowed, Demographics and Q8”,

proc import datafile="E:WORKNTOCT813313Borrowed.csv"

out=Borrowed

dbms=csv replace;

label q21a = "Borrowed in past 12 months: from a financial institution",

q21b = "Borrowed in past 12 months: from a store (store credit)",

q21c = "Borrowed in past 12 months: from family or friends",

q21d = "Borrowed in past 12 months: from another private lender",

q22a = "Borrowed in past 12 months: for education or school fees",

q22b = "Borrowed in past 12 months: for medical purposes",

q22c = "Borrowed in past 12 months: for farm/business purposes";

 run; 

 proc import datafile="E:WORKNTOCT813313Demographics.csv"

 out=Demographics

 dbms=csv replace;

 run; 

 proc import datafile="E:WORKNTOCT813313gov.csv"

 out=A3.government

 dbms=csv replace;

 run;

 proc import datafile="E:WORKNTOCT813313Q8.csv"

 out=Q8

 dbms=csv replace;

 run;

Now check the validity of the data, whether the data is primary or secondary, level of measurements and the analysis collection process. The SAS codes to calculate the frequency is given below:

proc freq data=A3.Borrowed; tables q21a q21b q21c q21d q22a q22b q22c; run;

The output of the above procedure is given below:

The FREQ Procedure

Borrowed in past 12 months: from a financial institution,
q21a	Frequency	Percent	Cumulative Frequency	Cumulative Percent
(dk)	2	0.20	2	0.20
(ref	2	0.20	4	0.40
no	804	80.24	808	80.64
yes	194	19.36	1002	100.00

Borrowed in past 12 months: from a store (store credit),
q21b	Frequency	Percent	Cumulative Frequency	Cumulative Percent
(dk)	4	0.40	4	0.40
no	894	89.22	898	89.62
yes	104	10.38	1002	100.00

Borrowed in past 12 months: from family or friends,
q21c	Frequency	Percent	Cumulative Frequency	Cumulative Percent
(dk)	1	0.10	1	0.10
no	885	88.32	886	88.42
yes	116	11.58	1002	100.00

Borrowed in past 12 months: from another private lender,
q21d	Frequency	Percent	Cumulative Frequency	Cumulative Percent
(d	3	0.30	3	0.30
no	991	98.90	994	99.20
ye	8	0.80	1002	100.00

Borrowed in past 12 months: for education or school fees,
q22a	Frequency	Percent	Cumulative Frequency	Cumulative Percent
(d	1	0.10	1	0.10
no	962	96.01	963	96.11
ye	39	3.89	1002	100.00

Borrowed in past 12 months: for medical purposes,
q22b	Frequency	Percent	Cumulative Frequency	Cumulative Percent
no	947	94.51	947	94.51
yes	55	5.49	1002	100.00

Borrowed in past 12 months: for farm/business purposes
q22c	Frequency	Percent	Cumulative Frequency	Cumulative Percent
no	971	96.91	971	96.91
ye	31	3.09	1002	100.00

Thus, there are total 1002 observations in the data, so data is very large. The data was collected from Word Bank, so it is a secondary data type.

The questions are answered in categories “yes and No”, so the variables in the data “Borrowed” are measured in nominal level of measurement.

Thus it can say that the analysis of the results will be reliable.

1. Now use data step to read the data, the SAS codes are given below:

 data A3.Borrowed;

 set Borrowed;

 run; 

 data A3.Demographics;

 set Demographics;

 run; 

 data A3.Q8;

 set Q8;

 run;replace;

 run;

2. To analyse of the EDUC level of participants, whether participants are primary or less, secondary and tertiary or above, create a new variable “EDUCATION” by assigning the variable names as “1=primary or less,2= secondary and 3= tertiary or above”. The codes are given below: 

data A3.Demographics;

 set Demographics;

if EDUC eq '1' then EDUCATION = "primary or less";

if EDUC eq '2' then EDUCATION = "Secondary";

if EDUC eq '3' then EDUCATION = "tertiary or above";

Results

run; 

To calculate the frequency of the each class, use the following codes: 

proc freq data=A3.Demographics; tables EDUCATION; run; 

The frequency of each participant by education is given below:

The FREQ Procedure

EDUCATION	Frequency	Percent	Cumulative Frequency	Cumulative Percent
Secondary	586	59.49	586	59.49
primary or less	51	5.18	637	64.67
tertiary or above	348	35.33	985	100.00

Frequency Missing = 17 

Thus, there are 586 participants are educated to secondary level, 51 participants are educated to primary or less and 348 participants are educated to tertiary or above level. 

3. The One-way ANOVA will be used to check whether there is a difference in the monthly income for new variable Education at 2% level of significance.  The one way analysis of variance is used to test whether there is significant relationship between the means of unrelated groups (Education) which have more than two levels (Maxwell & Riccardo, 2014). The SAS codes for the model “Income=education” is given as below:  

data A3.Demographics;

 set A3.Demographics;

income=input (month_inc,dollar11.);

run; 

PROC ANOVA data = A3.Demographics;

CLASS EDUCATION;

MODEL income = EDUCATION;

run; 

The results for the above procedure is given as below: 

The ANOVA Procedure

Class Level Information
Class	Levels	Values
EDUCATION	3	Secondary primary or less tertiary or above

Number of Observations Read	1002
Number of Observations Used	985

The ANOVA Procedure

Dependent Variable: income

Source	DF	Sum of Squares	Mean Square	F Value	Pr > F
Model	2	3221431512	1610715756	23.95	<.0001
Error	982	66034974151	67245391
Corrected Total	984	69256405663

R-Square	Coeff Var	Root MSE	income Mean
0.046515	80.65363	8200.329	10167.34

Source	DF	Anova SS	Mean Square	F Value	Pr > F
EDUCATION	2	3221431512	1610715756	23.95	<.0001

The value of R-square is 0.0465; so 4.65% of the variation in income can be described by education and rest of the variation remains unexplained. 

The calculated value of F-Statistic is 23.95. The P-value corresponding to numerator degree of freedom is 2 and denominator degree of freedom 982 is less than 0.0001. Now compare the P-value with 2% level of significance. The P-value (0.0001) is less than 2% level of significance, so the null hypothesis of the test gets rejected. 

Hence, the means for the three educations level is different corresponding to the income or participants. 

4. To compare the percentage of government assistance by each level of education use the following codes in SAS:

data A3.Government;

 set A3.Government;

label q39 = "Received government transfers in past 12 months";

run; 

proc sort data=A3.Demographics;

   by wpid_random;

run;

proc sort data=Government;

   by wpid_random;

run; 

data A3.Demo_Gover;

   merge A3.Government A3.Demographics;

   by  wpid_random;

run; 

proc freq data=A3.Demo_Gover; tables q39*EDUCATION; run; 

The obtained frequency is given below:

The FREQ Procedure

Frequency Percent Row Pct Col Pct

Table of Q39 by EDUCATION Q39(Received government transfers in past 12 months) EDUCATION Total Secondary primary or less tertiary or above yes 285 65.67 65.67 100.00 27 6.22 6.22 100.00 122 28.11 28.11 100.00 434 100.00     Total 285 65.67 27 6.22 122 28.11 434 100.00 Frequency Missing = 568

The government assistance for the secondary level education participants is highest. The government assistance for the secondary level education participants is 65.67%, for primary or less level is 6.22% and for tertiary or above level is 28.11%. 

5. To compare the percentage of Borrowed in past 12 months: from a financial institution and the  Borrowed in past 12 months: for education or school fees, use the following codes in SAS: 

proc freq data=Borrowed; tables q21a*q22a; run;

The results are shown below:

The FREQ Procedure

Frequency Percent Row Pct Col Pct

Table of q21a by q22a q21a(Borrowed in past 12 months: from a financial institution,) q22a(Borrowed in past 12 months: for education or school fees,) Total (d no ye (dk) 0 0.00 0.00 0.00 2 0.20 100.00 0.21 0 0.00 0.00 0.00 2 0.20     (ref 0 0.00 0.00 0.00 2 0.20 100.00 0.21 0 0.00 0.00 0.00 2 0.20     no 1 0.10 0.12 100.00 778 77.64 96.77 80.87 25 2.50 3.11 64.10 804 80.24     yes 0 0.00 0.00 0.00 180 17.96 92.78 18.71 14 1.40 7.22 35.90 194 19.36     Total 1 0.10 962 96.01 39 3.89 1002 100.00

The 19.36% of the participants Borrowed in past 12 months: from a financial institution for education or school fees and 80.24 % of the participants Borrowed in past 12 months: from a financial institution for other purpose. 

Out of total 1002 participants, 586 participants are educated to secondary level, 51 participants are educated to primary or less and 348 participants are educated to tertiary or above level. The government assistance for the secondary level education participants is 65.67%, for primary or less level is 6.22% and for tertiary or above level is 28.11%.  

The 19.36% of the participants Borrowed in past 12 months from a financial institution for education or school fees and 80.24 % of the participants Borrowed for other purpose.

Hence, government assistance for all education level is approximately equal as comparison to number of participates in each education level. And, the participants mostly borrowed for other purpose, so the borrowing habits are different for the all education levels.

Conclusion 

There are 586 participants are educated to secondary level, 51 participants are educated to primary or less and 348 participants are educated to tertiary or above level.

The value of R-square is 0.0465; so 4.65% of the variation in income can be described by education and rest of the variation remains unexplained. The three educations level is different corresponding to the income or participants. 

The government assistance for the secondary level education participants is highest. The government assistance for the secondary level education participants is 65.67%, for primary or less level is 6.22% and for tertiary or above level is 28.11%.

The 19.36% of the participants Borrowed in past 12 months: from a financial institution for education or school fees and 80.24 % of the participants Borrowed in past 12 months: from a financial institution for other purpose. 

References

Maxwell, R., & Riccardo, R. (2014). A Student's Guide to Analysis of Variance.