Scribd
Upload
77 views4 pages

Regression Analysis and Its Interpretation With Spss

This document provides details on regression analysis conducted to model the relationship between household consumption (C), income (Y), and family size (F). The regression model found that 85.9% of the variation in household consumption (C) can be explained by income (Y) and family size (F). Income and family size were both found to have a statistically significant positive relationship with consumption. Specifically, a 1% increase in income increases consumption by 0.842%, and a 1% increase in family size increases consumption by 0.216%. Therefore, higher income levels and larger family sizes contribute to increased household consumption.

Uploaded by

ChakraKhadka
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
77 views4 pages

Regression Analysis and Its Interpretation With Spss

This document provides details on regression analysis conducted to model the relationship between household consumption (C), income (Y), and family size (F). The regression model found that 85.9% of the variation in household consumption (C) can be explained by income (Y) and family size (F). Income and family size were both found to have a statistically significant positive relationship with consumption. Specifically, a 1% increase in income increases consumption by 0.842%, and a 1% increase in family size increases consumption by 0.216%. Therefore, higher income levels and larger family sizes contribute to increased household consumption.

Uploaded by

ChakraKhadka
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Research Methodology Hands Out for BHTM VII Semester: Regression Analysis and its 2013,

interpretation with SPSS; CHAKRA B. KHADKA, PhD LA College December

REGRESSION ANALYSIS AND ITS INTERPRETATION WITH


SPSS

Consumption Income Family size lnC lnI lnF


4.60 24.00 1.00 3.66276 4.38021 0
5.10 40.00 1.00 3.70757 4.60206 0
5.20 28.00 1.00 3.716 4.44716 0
6.40 47.00 1.00 3.80618 4.6721 0
5.60 32.00 2.00 3.74819 4.50515 0.30103
8.10 59.00 2.00 3.90849 4.77085 0.30103
5.80 30.00 2.00 3.76343 4.47712 0.30103
5.10 31.00 2.00 3.70757 4.49136 0.30103
2.90 26.00 2.00 3.4624 4.41497 0.30103
5.20 28.00 3.00 3.716 4.44716 0.47712
5.10 26.00 3.00 3.70757 4.41497 0.47712
7.80 30.00 3.00 3.89209 4.47712 0.47712
4.90 42.00 3.00 3.6902 4.62325 0.47712
7.90 42.00 3.00 3.89763 4.62325 0.47712
11.30 54.00 4.00 4.11394 4.73239 0.60206
11.80 58.00 4.00 4.07188 4.76343 0.60206
4.80 20.00 5.00 3.68124 4.30103 0.69897
13.70 85.00 5.00 4.13672 4.92942 0.69897
18.00 82.00 6.00 4.25527 4.91381 0.77815
20.00 112.00 6.00 4.30103 5.04922 0.77815

Data description: C = households' annual consumption expenditure in Rs '000;


Y=households' annual income (in Rs '000); F= number of family; ln=log to the base 10
(The data are transformed into logarithmic form to smooth out them and also to make the
interpretation of the result easier).

Result of the Ordinary least Square (OLS) Estimate of the equation


lnC = a + blnY+ dlnF (a, b and d are the constants and parameters to be estimated) in
SPSS is presented below:

1|Page
Research Methodology Hands Out for BHTM VII Semester: Regression Analysis and its 2013,
interpretation with SPSS; CHAKRA B. KHADKA, PhD LA College December

Table 1: Variables Entered/Removed (b)

Model Variables Entered Variables Removed Method


LNF, LNY(a) . Enter
a. All requested variables entered.
b. Dependent Variable: LNC
This table just provides information on dependent and independent variables interred or
removed in the regression equation.

Table 2: Model Summary (b)


Adjusted R Std. Error of the
Model R R Square Square Estimate Durbin-Watson
1 .935(a) .874 .859 .08182 1.870
A. Predictors: (Constant), LNF, LNY
b. Dependent Variable: LNC

This table provides important information on some aspect of the estimated result; as it is
clear the highlighted parameters are R square, adjusted R square, standard error of the
estimate and Durbin-Watson static. R square or adjusted R informs the explained
variation in the dependent variable by the estimated regression line. The value of adjusted
R squared is .859 which means that around 85.9 % variation in the consumption
expenditure is explained by the regression equation involving two explanatory variables
income and number of family. Adjusted R square is preferred to simple R square because
Adjusted R square considers the effect of added more variables.

The Durbin-Watson statistic is the statistic which is used to detect whether there is any
serial or autocorrelation between the residual terms, residual = actual value of the
dependent variable - estimated value of the dependent variable. The tolerable value of
Durbin-Watson statistic is between 1.5 to 2.4.In the estimated equation the value of
Durbin-Watson statistic is 1. 87. The value of DW indicates that there is no serial
autocorrelations between the residual terms.

2|Page
Research Methodology Hands Out for BHTM VII Semester: Regression Analysis and its 2013,
interpretation with SPSS; CHAKRA B. KHADKA, PhD LA College December

Table 3: ANOVA (b)

Model Sum of Squares df Mean Square F Sig.


1 Regression .789 2 .395 58.949 .000(a)
Residual .114 17 .007
Total .903 19
a .Predictors: (Constant), LNF, LNY
b. Dependent Variable: LNC

Table 3 brings information on the explained, unexplained and total variation in the
dependent variable. Regression sum of square (RSS) which is also called explained
sum of square (ESS) is .789 and this is the sum of square explained by the fitted
regression line. The other is the Residual sum of squares or error sum of squares, an
unexplained sum of squares which is .114.Mean square of the regression is .395 and
mean square of the residual sum of square is .007. These are used to compute F statistic.F
value is considerably high, 58.949, which is significant at more than 99 % confidence
level (The probability of F- statistic is .000 which is less than 0.05 or 0.01 % level of
significance). Generally higher the F -statistic the better is the fit of the regression
equation.

Table 4: Coefficients (a)


Unstandardized Standardized
Model Coefficients Coefficients t Sig.
Variables B Std. Error Beta
1 (Constant) -.599 .159 -3.765 .002
LNY .842 .107 .782 7.886 .000
LNF .216 .084 .256 2.580 .019
a. Dependent Variable: LNC

Table 4 is the most important part of the regression result. It gives the estimate of the
regression equation lnC = a + blnY + dlnF. Generally the unstandardized coefficients

3|Page
Research Methodology Hands Out for BHTM VII Semester: Regression Analysis and its 2013,
interpretation with SPSS; CHAKRA B. KHADKA, PhD LA College December

are taken the estimated coefficients of the equation. The estimated model is written as:
lnC = - 0. 599 + 0.842 lnY + 0. 216 lnF. The coefficient of lnY is positive with a value
of 0.842 which means that if income increases by 1% consumption expenditure increases
by 0.842 %, that is if, income increases by Rs100, consumption increase by Rs84.2, so
the marginal propensity to consume of the sampled households is 0.842.Similarly the
coefficient of lnF is also positive and its value is 0.216 which means that if the number of
families increase by 1% consumption expenditure increases by 0.216 % on the average.
Increased population needs food, shelter and cloth and the expenses increase definitely.
The t-statistic indicates whether the unstandardised or the standardized coefficients are
statistically significant or not. The rule of thumb is that if t-statistic is 2 (two) the
coefficients are statistically significant at 95 % confidence level (5 % significance level
i.e, at 5% chance of error).On this rule , the coefficients of the constant term (-.599), of
lnY (.842) and of lnF (.216) all are statistically significant .The positively and
statistically significant coefficient of lnY and lnF mean that household's income level
and family/household sizes are the factor that contribute in the increase of consumption
expenditure.

4|Page

You might also like