Question

2. “Question2_Dataset.xlsx” contains the following data for several underdeveloped countries:

• Infant mortality rate
• Adult literacy rate
• Percentage of students finishing primary school
• Per capita GNP
  1. Use this data to develop an equation that can be used to predict infant mortality.
  2. Are there any outliers in this set of data?
  3. Interpret the coefficients in your equation. (What does each coefficient number mean?)
  4. Within what value should 95% of your predictions be infant mortality be accurate?

	Infant Mortality(deaths per thousand births)	%age adult literacy	%age finishing primary school	GNP per capita	Predictions	Residuals	Residuals^2
Cuba	18	98	98	2000	0	18	324
Sri Lanka	20	85	92	3300	0	20	400
Costa Rica	19	94	84	5800	0	19	361
Vietnam	44	85	58	600	0	44	1936
China	54	80	86	2400	0	54	2916
South Africa	56	76	68	4000	0	56	3136
Saudi Arabia	38	59	68	11000	0	38	1444
Brazil	60	78	56	5600	0	60	3600
Zimbawe	68	82	76	1800	0	68	4624
Morocco	68	42	76	3400	0	68	4624
Pakistan	98	36	38	2100	0	98	9604
Nigeria	86	44	56	1600	0	86	7396

Answer 1

Using Excel

data -> data analysis -> regression

SUMMARY OUTPUT
Regression Statistics
Multiple R	0.919743
R Square	0.845926
Adjusted R Square	0.788149
Standard Error	11.93399
Observations	12
ANOVA
	df	SS	MS	F	Significance F
Regression	3	6255.555	2085.185	14.64108	0.001298
Residual	8	1139.361	142.4202
Total	11	7394.917
	Coefficients	Standard Error	t Stat	P-value	Lower 95%	Upper 95%
Intercept	154.2055	16.15218	9.547035	1.2E-05	116.9585	191.4525
%age adult literacy	-0.67719	0.228723	-2.96074	0.018126	-1.20463	-0.14975
%age finishing primary school	-0.5982	0.275808	-2.16889	0.061925	-1.23421	0.037816
GNP per capita	-0.00293	0.00129	-2.27093	0.052813	-0.0059	4.52E-05

Infant mortality rate^ = 154.2055 -0.67719 % age adult literacy -0.5982 %age finishing primary school - 0.00293 GNP per capita

slope is interpreted as when we change independent variable(x) by 1 unit, on average dependent variable(y) will change by slope units

hence when % age adult literacy increase by 1, on average Infant mortality rate decreases by 0.67719 unit,

similarly others

standard error = 11.93399

this means on average predicted value will differ from actual value by 11.934 unit