Statistical Analysis of Business Startup Costs

This assignment focuses on analyzing the startup costs of different businesses using statistical methods. Task 1 involves calculating descriptive statistics like mean, median, and mode, creating frequency distribution tables and histograms to visualize the data. ANOVA is used to test for significant differences in startup costs across businesses. Task 2 utilizes regression analysis to model the relationship between startup costs and other factors. The assignment assesses the goodness of fit of the model, examines the significance of predictor variables, and interprets the results using R-squared values, p-values, and confidence intervals.

STATISTICS
HI6007
[Pick the date]
Student id

Task 1
1. The required tables for illustrating the excel output in relation to key descriptive statistics are
illustrated.
2.
(a) Frequency Distribution Tables
1

(b) Histograms (Frequency Distribution Data)
2

3

4

3) Observation 1: Startup costs across business show variation which may be significant.
Reason: Difference in the central tendency measures of different businesses, Histograms having
different distribution despite same class size.
Observation 2: Startup cost distribution is not normal
Reason: Skew presence is indicated from the unsymmetrical histogram shape. Also, the
difference found in mean and median also reflect this.
4)
The test to be used is ANOVA and not a t-test owing to variables undergoing mean comparison
is five (hence higher than 2).
5

Decision: H0 rejected as 0.02 (p value) < 0.05 (significance level)
Result: Significant variation observed across startup costs in different businesses.
TASK 2
1) MS-Excel Regression Output
6

Regression Equation
2) Objective : To comment on fit
Indicator to be used: Coefficient of Determination or R2
Observation: R2 = 0.9932
Conclusion: The model is a good fit considering high predictive power of the regression model.
3)
EXCEL OUTPUT
Observation: Significance Level (0.05) > Significance F(0.0000)
Decision: Rejection of H0
Conclusion: Model is significance as there exists a significant slope which is non-zero.
4) Interpretation Table
5) Table for highlighting confidence interval
7

6)
The p values for the individual variable slope are estimated from the output highlighted below.
Observation: Significance Level > P value for all predictors
Result: Rejection of H0 for every predictor
Conclusion: The significance of slope of each of the predictor (independent) variables is
established.
7) Owing to significance testing of the various predictors above, no alternation in the model is
recommended considering the fact that all variables have a significant slope.
8) The relevant computations are as follows.
8

9