代写Problem set 11 Final Exam - KEY代做Prolog

Final Exam - KEY

1.    A study looked at the impact of whether or not a first year student attended orientation and if they lived on campus or not on their overall satisfaction with student services (measured on a  ten point scale at the end of the first year).  A two by two factorial analysis of variance was performed via multiple regression.  The results indicated a significant interaction.  Using the means table below, plot the interaction and interpret the results.

Commute

The graph of the significant interaction term indicates the effect of commuting on satisfaction

depends on the whether or not one attended orientation. The effect of commuting was stronger for those who did not attend orientation (there is a sharper decrease in satisfaction). There is little

difference in satisfaction for those who do not commute regardless of whether or not they attended orientation.  However, there is a larger difference who commuted. Of those who commuted, those who attended orientation reported higher satisfaction.  A significance testis required to determine if the size of difference in satisfaction, is indeed statistically significant.

2.    Given the data and contrast codes below, write the estimated regression model Y(̂)i =

b0 +b1C1+b2C2

Descriptive statistics:

3.    A highschool introduced an experiential learning component to their science courses in 11th grade.  In the past, all students participated in an in-school, applied project over the course of the school year.  This year, instead, they randomly assigned students to different programs/projects in and outside the school.  Group 1 was assigned to the traditional, existing in-school applied project.  Group 2 was assigned to participate in an engineering program at a  local firm, Group 3 was assigned to a medical research center and group 4 was assigned to a  children’s science museum.  At the end of the school year, all students took an end of year assessment, aligned to the curriculum,  to determine their level of scientific skills and knowledge.  The scale scores range from 0 to 20.  To analyze the data and compare performance by group, the following codes (see below) were used.  Using the information below and the SPSS output, summarize the results of the analysis.  Be concise but include all relevant statistical evidence.

The difference in experiential setting explains approximately 37.1% of the variance in the assessment scores.  Setting is a significant predictor of assessment score since F = 8.664 with p < .05. Since nearly 63% of the variance is unexplained there are likely other important predictors of the science assessment score missing.  The estimated regression model is Y(̂)i =   14. 083 + 2. 833D1 − 1. 083D2 +

2. 25D3 .  D1 is a significant predictor of assessment score since the t-value of 3.188 has ap-value of

approximately .003.  This indicates there is a significant difference between the assessment scores of groups 1 (traditional) and 2 (engineering program). Students who attended the engineering program scored approximately 2.833 points higher than those in the traditional project. There was not a

significant difference between the traditional program and the medical research center (t=-1.219, p<.05).   Finally, students who participated in the children’s science museum scored 2.532 points higher on the assessment.  This is a significant amount since t = 2.532 with p = .015.

Report

Y

Group	Mean	N	Std. Deviation
1.00	14.0833	12	2.31432
2.00	16.9167	12	2.27470
3.00	13.0000	12	1.90693
4.00	16.3333	12	2.18812
Total	15.0833	48	2.65645

Model Summary

Model	R	R Square	Adjusted R Square	Std. Error of the Estimate
1	.609a	.371	.328	2.17684

a. Predictors: (Constant), D3, D2, D1

ANOVAa

Model	Sum of Squares	df	Mean Square	F	Sig.
1 Regression Residual Total	123.167	3	41.056	8.664	.000b
	208.500	44	4.739
	331.667	47

a. Dependent Variable: Y

b. Predictors: (Constant), D3, D2, D1

Coefficientsa

	Unstandardized		Standardized
Model	Coefficients		Coefficients	t	Sig.
	B	Std. Error	Beta
1 (Constant) D1 D2 D3	14.083	.628		22.411	.000
	2.833	.889	.467	3.188	.003
	-1.083	.889	-.178	-1.219	.229
	2.250	.889	.371	2.532	.015

a. Dependent Variable: Y

4.    An organization provides training courses in the use of their class scheduling software for

schools. This particular course is delivered to administrators (participants) who had never previously used software to schedule classes. An end of course project is given to all participants and is graded out of 100 points. The organization would like to test whether mode of delivery    (online (coded -1) or in person (coded 1)) results in different performance on the course ending   project.  To control for differences in computer knowledge, the organizers administer a computer software literacy test (50-point scale) prior to the training.  Participants are randomly assigned to the two modes of delivery (18 to each mode).  They conduct an analysis of covariance to determine if the different modes result in different scores on the end of course assessment.  Use the output on the following pages to summarize the findings.  Be concise while including the relevant statistical evidence.

As this is an analysis of covariance, the assumption of homogeneity of slopes should first be checked. Since the interaction term is not significant (t=-1.897, P >.05), the assumption is not violated.

Software literacy appears to be a strong covariate (though we do not have significant tests) as it is correlated with project score (r=.649) and not strongly correlated with mode (r = .289, from tolerance values). Software literacy and mode of instruction together significantly (F = 133.081, P < .05) predict the end of course project.  Together they explain 89% of the variance in scores. Both software literacy (t=14.182, p < .05) and mode (t=-11.833, p < .05) are significant predictors of score controlling for the other.  When controlling for software literacy, participants that participate in online training scored approximately 6.474 (2 X 3.237) points higher than those attending in person.