discussion 6 (1)

ANALYTICS REPORT TO: FUTURE EMPLOYER FROM: TATUM SANUIK SUBJECT: NON-LINEAR REGRESSION DATE: NOVEMBER 5, 2023 Introduction . In order to improve goods and services, this document examines data from recent consumer satisfaction surveys. It pinpoints areas for improvement, provides guidance for corporate strategy, and focuses on client satisfaction. The report, which is meant for executives and managers, provides data-driven guidance for better client loyalty and corporate expansion. To improve overall satisfaction, promptly address the most important survey issues and conduct focused training and product enhancements in line with customer input. It is clear from the data that the Log-Log model is the most appropriate model to use in this situation. Data Analysis Quadratic Model ^ MCMS = -1082.44 + 1.70(SMS) + 2.61E-06 ( SMS ) 2 Marginal Effect: b 1 + 2b 2 x = 1.70 + 2(2.61E-06)(54000) = 1.98 As starting salary increases by $1 from $54,000 to $54,001, mid-career salary increases by $1.98 on average and all else constant. The shape of this quadratic would be concave-up because the coefficient for the squared term is positive. Lin-Log Model ^ MCMS = -855271.75 + 874.06ln(SMS) + ( SMS ) 2 Marginal Effect: : b 1 + 2b 2 x = 874.06 + 2 ( SMS ) 2 ( 54000 ) = 874.06 As starting salary increases by 1%, mid-career salary increases by $874.06 on average and all else constant. Log-Lin Model ^ MCMS = 10.22+ 874.06ln(SMS) + ( SMS ) 2 Marginal Effect: b 1 + 2b 2 x = 2.38 + 2 ( SMS ) 2 ¿ (54000) = 2.4 As starting salary increases by $1,000, mid-career salary increases by 2.4%, on average and all else constant. Page 1 of 4

Log-Log Model ^ MCMS = -0.27+ 874.06ln(SMS) + ( SMS ) 2 Marginal Effect: b 1 + 2b 2 x = 1.0 + 2 ( SMS ) 2 ¿ (54000) = 0.058 As starting salary increases by $1, mid-career salary increases by 0.058, on average and all else constant. Best Model ^ MCMS = e − 0.27 + 1.08ln ( 54000 ) + ¿ 0.058 2 2 ¿ Log-Log is the best model to use. Predictions of ln(MCMS) using this model are off by an average of 0.058 log units. Conclusion Recent customer satisfaction survey data has been analyzed to identify particular areas for improvement. This has made it possible to develop well-informed company plans that are intended to increase customer satisfaction and loyalty. It is advised to take immediate action to resolve the survey flaws that were found, as well as to introduce focused training and product enhancements that are in line with user input. It is advised for this case to use the Log-Log model, which offers a useful framework for handling customer satisfaction issues, based on the data analysis. Page 2 of 4

SUMMARY OUTPUT Regression Statistics Multiple R 0.868361615 R Square 0.754051894 Adjusted R Square 0.752343922 Standard Error 0.058100941 Observations 146 ANOVA df SS MS F Significance F Regression 1 1.490344203 1.490344203 441.4893645 1.05456E-45 Residual 144 0.486103591 0.003375719 Total 145 1.976447794 Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Intercept 10.21985415 0.05068339 201.6410935 1.7986E-178 10.11967462 10.32003 Starting Median Salary 2.38545E-05 1.1353E-06 21.01164831 1.05456E-45 2.16105E-05 2.61E-05 Tatum Sanuik SUMMARY OUTPUT Regression Statistics Multiple R 0.869571893 R Square 0.756155278 Adjusted R Square 0.754461912 Standard Error 0.057851964 Observations 146 ANOVA df SS MS F Significance F Regression 1 1.494501431 1.494501431 446.5397447 5.67423E-46 Residual 144 0.481946363 0.00334685 Total 145 1.976447794 Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Intercept -0.27249378 0.546716328 -0.498418953 0.618949213 -1.353119642 0.808132083 ln(SMS) 1.079923359 0.051104939 21.13148704 5.67423E-46 0.978910611 1.180936108 Tatum Sanuik MCMS as y-variable LN(MCMS) a y- Quadratic 0.76665412 Log-Lin 0.75234392 Lin-Log 0.76190219 Log-Log 0.75446191 Quadratic 0.76665412 Log-Log 0.75446191 0.76829518 Tatum Sanuik Page 4 of 4