Regress internet usage on survey year only (i.e., predict a person's internet usage by survey year).

• Show the EXCEL or SPSS output of this regression.

SUMMARY OUTPUT
Regression Statistics
Multiple R	0.213777
R Square	0.045701
Adjusted R Square	0.045615
Standard Error	13.18691
Observations	11165
ANOVA
	df	SS	MS	F	Significance F
Regression	1	92961.87	92961.87	534.5875	1.4E-115
Residual	11163	1941185	173.8946
Total	11164	2034147
	Coefficients	Standard Error	t Stat	P-value	Lower 95%	Upper 95%	Lower 95.0%	Upper 95.0%
Intercept	5.414155	0.216203	25.04202	1.1E-134	4.990359	5.83795	4.990359	5.83795
year	0.460392	0.019912	23.12115	1.4E-115	0.42136	0.499423	0.42136	0.499423

• How does internet usage change when survey year increases by 1?
• Is the coefficient you described in b significantly different from 0?
• If the survey year was 2000 (i.e., year = 0), what would we expect the respondent's internet usage to be?
• What proportion of the variance in internet usage does this year explain?
• Does survey year account for a significant proportion of the variance in internet usage?

Regress internet usage on marital status (i.e., predict a person's internet usage using whether they are married or not).

• Show the EXCEL or SPSS output of this regression.

SUMMARY OUTPUT
Regression Statistics
Multiple R	0.082201
R Square	0.006757
Adjusted R Square	0.006668
Standard Error	13.45329
Observations	11165
ANOVA
	df	SS	MS	F	Significance F
Regression	1	13744.59	13744.59	75.94072	3.34E-18
Residual	11163	2020402	180.991
Total	11164	2034147
	Coefficients	Standard Error	t Stat	P-value	Lower 95%	Upper 95%	Lower 95.0%	Upper 95.0%
Intercept	10.58233	0.178178	59.39206	0	10.23307	10.93159	10.23307	10.93159
married	-2.21955	0.254699	-8.7144	3.34E-18	-2.7188	-1.72029	-2.7188	-1.72029

• What is the predicted difference in internet usage between married and unmarried respondents?
• Is the coefficient you described in b significantly different from 0?
• If the respondent were unmarried, what would we expect their internet usage to be?
• What proportion of the variance in internet usage does sex explain?
• Does marital status account for a significant proportion of the variance in internet usage?

Regress internet usage on whether the individual is currently employed or not (i.e., predict internet usage using whether the individual was employed or not).

• Show the EXCEL or SPSS output of this regression.

SUMMARY OUTPUT
Regression Statistics
Multiple R	0.038749
R Square	0.001502
Adjusted R Square	0.001412
Standard Error	13.48883
Observations	11165
ANOVA
	df	SS	MS	F	Significance F
Regression	1	3054.304	3054.304	16.78663	4.21E-05
Residual	11163	2031093	181.9486
Total	11164	2034147
	Coefficients	Standard Error	t Stat	P-value	Lower 95%	Upper 95%	Lower 95.0%	Upper 95.0%
Intercept	8.711671	0.230117	37.85765	3.6E-295	8.260602	9.16274	8.260602	9.16274
working	1.133176	0.276577	4.097149	4.21E-05	0.591037	1.675315	0.591037	1.675315

• What is the predicted difference in internet usage when the respondent is currently employed rather than unemployed?
• Is the coefficient you described in b significantly different from 0?
• What would we expect internet usage to be for a respondent not currently employed?
• What proportion of the variance in internet usage is explained in this model?
• Does employment status account for a significant proportion of the variance in internet usage?

Regress internet usage on educational attainment, using less than high school as the reference category.

• Show the EXCEL or SPSS output of this regression

SUMMARY OUTPUT
Regression Statistics
Multiple R	0.049169
R Square	0.002418
Adjusted R Square	0.002149
Standard Error	13.48385
Observations	11165
ANOVA
	df	SS	MS	F	Significance F
Regression	3	4917.823	1639.274	9.016204	5.84E-06
Residual	11161	2029229	181.8143
Total	11164	2034147
	Coefficients	Standard Error	t Stat	P-value	Lower 95%	Upper 95%	Lower 95.0%	Upper 95.0%
Intercept	8.067541	0.522097	15.45219	2.6E-53	7.044139	9.090943	7.044139	9.090943
degree_hs	0.989851	0.552904	1.790277	0.073436	-0.09394	2.07364	-0.09394	2.07364
degree_jc	1.670508	0.668135	2.500258	0.012425	0.360847	2.98017	0.360847	2.98017
degree_coll	2.213121	0.564336	3.92164	8.85E-05	1.106924	3.319318	1.106924	3.319318

• What is the predicted difference in internet usage when respondent has a college degree rather than a less than high school education?
• Is the coefficient you described in b significant?
• What is the internet usage we would expect for someone with less than high school education?
• What proportion of the variance in internet usage does respondent's education explain?
• Does the respondent's education account for a significant proportion of the variance in internet usage?

Regress internet usage on respondent's education (less than high school is the reference), survey year, size of respondent's city, number of children in the house, sex, marital status, and employment status.

• Show the EXCEL or SPSS output of this regression

SUMMARY OUTPUT
Regression Statistics
Multiple R	0.247351
R Square	0.061183
Adjusted R Square	0.060425
Standard Error	13.08419
Observations	11165
ANOVA
	df	SS	MS	F	Significance F
Regression	9	124454.3	13828.26	80.77435	7.5E-146
Residual	11155	1909693	171.1961
Total	11164	2034147
	Coefficients	Standard Error	t Stat	P-value	Lower 95%	Upper 95%	Lower 95.0%	Upper 95.0%
Intercept	2.299921	0.601306	3.824877	0.000132	1.121255	3.478586	1.121255	3.478586
degree_hs	1.816688	0.540514	3.361037	0.000779	0.757185	2.876191	0.757185	2.876191
degree_jc	2.745874	0.654431	4.19582	2.74E-05	1.463074	4.028674	1.463074	4.028674
degree_coll	3.298204	0.556417	5.927579	3.17E-09	2.207529	4.388878	2.207529	4.388878
year	0.472416	0.019994	23.62828	1.7E-120	0.433225	0.511607	0.433225	0.511607
size	0.244671	0.109528	2.233862	0.025512	0.029976	0.459366	0.029976	0.459366
kid	0.039239	0.125163	0.313504	0.753904	-0.2061	0.28458	-0.2061	0.28458
male	1.338612	0.252086	5.310145	1.12E-07	0.844479	1.832745	0.844479	1.832745
married	-2.04469	0.255852	-7.99169	1.46E-15	-2.5462	-1.54317	-2.5462	-1.54317
working	1.417533	0.273431	5.184247	2.21E-07	0.88156	1.953506	0.88156	1.953506

• How does internet usage change for every additional million residents of the respondent's city, all else held constant?
• How does internet usage change when the respondent is male rather than female, all else held constant?
• How does internet usage change when the respondent is married rather than unmarried, all else held constant?
• How does internet usage change when respondent has a high school diploma rather than less than high school education, all else held constant?
• What does the intercept from this model mean conceptually?
• What proportion of the variance of internet usage is accounted for by these variables together?
• Does the model account for a significant proportion of the variance in internet usage?