CASE STUDY SCENERIO Hypothesis Testing Transcript This document provides a complete text based transcript of the interactive tutorial Introduction From Data to Decision Hypothesis testing is the structured process of using data to validate or reject an assumption about the real world It's how we move from I think to I can prove and make decisions based on evidence, not just intuition Core Concepts The Null Hypothesis (H) This is the default assumption, the status quo It states there is no effect or no difference For example, The new training method has no effect on technician speed Your goal is to find enough evidence to reject this claim The Alternative Hypothesis (H) This is what you are trying to prove It states there is an effect or a difference For example, The new training method does improve technician speed The p value This is the most critical output The p value is the probability of observing your data (or something more extreme) if the null hypothesis were true A small p value (typically 1 The t test Comparing Two Groups Description Use a t test when you want to compare the average (mean) of two, and only two, groups It tells you if the difference between their means is statistically significant or just due to random chance Apex Scenario Training Method Effectiveness Question Is our new simulation based training better than the traditional method H Mean completion time is the same for both groups H Mean completion time is different for the two groups Chart Description A bar chart compares the Avg Completion Time (Hours) for two groups Traditional Training 31 hours New Simulation 26 hours The tutorial notes that if the p value for this comparison is low ( 2 ANOVA Comparing More Than Two Groups Description ANOVA (Analysis of Variance) is an extension of the t test Use it when you need to compare the means of three or more groups It checks if there's a significant difference somewhere among the groups by comparing the variation between groups to the variation within each group Apex Scenario Maintenance Provider Effect Question Do different maintenance providers (QuickTurn, Premier, In house) have different impacts on reliability H The mean reliability is the same across all providers H At least one provider has a different mean reliability Chart Description A bar chart compares the Mean Time Between Failures (Hours) for three groups QuickTurn MRO 3,847 hours In House 6,000 hours Premier MRO 7,200 hours The tutorial notes that a low p value from ANOVA tells you that at least one group is different , but not which one You then use post hoc tests (like Tukey's HSD) to find the specific differences 3 The Chi Square () Test Analyzing Categories Description Use this test when working with categorical data (names, types, yes no) It determines if there is a significant association between two categorical variables by comparing the observed frequencies in your data to the frequencies you would expect if there were no relationship Apex Scenario Bay Configuration Impact Question Is there a relationship between the Maintenance Bay used and the occurrence of Rework H Bay assignment and rework are independent H Bay assignment and rework are associated Table Description A table shows the Observed Frequencies of rework Bay 1 (New) 5 Reworks, 95 No Reworks Bay 2 (Old) 15 Reworks, 85 No Reworks The tutorial notes that the Chi Square test calculates if the observed count of 5 reworks in Bay 1 is significantly lower than what you'd expect by random chance 4 Regression Analysis Finding Relationships Predicting Outcomes Description Regression analysis goes beyond simple comparisons It models the relationship between a dependent variable (the outcome you want to predict) and one or more independent variables (the drivers) It's incredibly powerful for forecasting and understanding the why behind your data Apex Scenario Turbine Blade Failure Question What drives engine failure Model Failure Rate (Blade Age) (Cycles) Interpretation The model tells us exactly how much the failure rate increases (the coefficient ) for each additional year of blade age, while controlling for the effect of cycles Chart Description A scatter plot shows Engine Failure Data with Blade Age (Years) on the x axis and Failure Rate ( ) on the y axis The points show a clear upward trend A Regression Line is drawn through the data, visually representing the relationship modeled by the formula The tutorial notes that the R squared value (e g , 0 71) tells you the percentage of variation in the outcome that is explained by the model's inputs Apex Aviation The Investigation Deepens Transcript This document provides a complete text based transcript of the interactive presentation Slide 1 The Investigation Deepens We need to know WHY What's driving these failures Maria Rodriguez Slide 2 The Task Force Reconvenes of Data The Horizon Task Force reconvenes in AAS' engineering lab The walls are now covered with scatter plots and Pareto charts Marcus Thompson stands before a massive display showing component failure data Maria Rodriguez opens Last week's analysis was sobering We have a bimodal fleet with four aircraft in critical condition Now we need to know WHY What's driving these failures Marcus begins I've spent the week testing three hypotheses Turbine blade age is the primary failure driver QuickTurn MRO's maintenance quality is substandard Cargo conversion stress accelerates deterioration Let me show you what the data reveals Slide 3 The Interrogation of Data Marcus Thompson Tests three core hypotheses to find the root cause of failures Quote I tested whether turbine blade age predicts engine failures James Chen Identifies critical patterns in the data that others might miss Quote Look at this clustering Ahmed Hassan Questions the data, ensuring all variables are considered Quote But is it just age, or is it cycles Lisa Park Focuses on external factors and accountability of suppliers Quote Can we prove QuickTurn is actually worse, not just unlucky Sarah Williams Challenges assumptions to prevent costly misinterpretations of the data Quote Maybe QuickTurn gets the worst aircraft Slide 4 Hypothesis 1 The Turbine Blade Investigation Heading Hypothesis 1 The Turbine Blade Investigation Question Is blade age the primary driver of engine failures Section 1 The Statistical Proof Test Result (t test) p value Conclusion The effect of blade age on failures is statistically significant Regression Analysis R 0 71 Factoid R squared (R) is a statistical measure that represents the proportion of the variance for a dependent variable that's explained by an independent variable An R of 0 71 means that 71 of the variation in engine failures can be explained by blade age and cycles combined Dialogue Age is the stronger predictor, but cycles matter too Together they explain 71 of the variance in failures Marcus Section 2 Interaction Dialogue Look at this clustering every engine failure above 50,000 cycles had blades over 10 years old James Dialogue But is it just age, or is it cycles We need to be sure Ahmed Slide 5 Hypothesis 2 The Maintenance Provider Effect Heading Hypothesis 2 The Maintenance Provider Effect Question Is QuickTurn MRO's maintenance quality substandard Dialogue Can we prove QuickTurn is actually worse, not just unlucky Lisa Section 1 The Verdict Finding ANOVA results show QuickTurn's maintenance leads to a significantly lower MTBF (p Factoid ANOVA (Analysis of Variance) is a statistical test used to determine whether there are any statistically significant differences between the means of two or more independent groups A p value Challenge Maybe QuickTurn gets the worst aircraft Sarah Response Marcus's matched pairs analysis controlled for this, confirming the provider effect was real and not just due to them servicing older planes Slide 6 Hypothesis 3 Cargo Conversion Stress Heading Hypothesis 3 Environmental Factors Question Does cargo conversion accelerate deterioration Section 1 Initial Analysis H Cargo conversion timing has no effect on reliability H Earlier conversions lead to worse reliability MTBF 8234 892 Years Since Conversion R 0 23, p 0 08 The relationship exists but isn't as strong as expected Section 2 Interaction Model Watch what happens when I add an interaction term MTBF 9012 234 Years Since Conversion 156 Aircraft Age 88 Years Since Conversion Aircraft Age R 0 58, p Ahmed Hassan interprets So older aircraft suffer more from cargo conversion stress The 2021 conversion on 23 year old aircraft were a perfect storm Key Insight The cargo conversion effect is not linear it compounds with aircraft age, creating exponentially worse reliability in older converted aircraft Slide 7 The Smoking Gun Component Interaction Analysis Logistic Regression Model logit(P) 4 23 1 34OldBlades 0 89QuickTurn 0 67HighCycles 2 45(OldBladesQuickTurn) When old blades QuickTurn maintenance Failure probability increases 11 6x 95 CI 6 2x to 21 7x James Chen stands up That's it QuickTurn isn't properly inspecting turbine blades They're missing the early warning signs Slide 8 The Predictive Model Using all significant factors to predict failures Random Forest Model Results Overall Accuracy 89 Sensitivity 84 Specificity 91 Hour Prediction 1,000 Variable Importance Blade age 28 Last maintenance provider 22 Cycles since conversion 18 Interaction terms 15 High Risk Aircraft (Next 1,000 Hours) N411HC 87 failure probability N416HC 84 failure probability N417HC 79 failure probability N404HC 72 failure probability N412HC 71 failure probability I can now predict with 89 accuracy which aircraft will have major failures in the next 1,000 hours Marcus Thompson The investigation has successfully identified the primary drivers of unreliability in the Horizon fleet Primary Driver Substandard maintenance from QuickTurn MRO is the single largest contributor to poor reliability, independent of aircraft age Secondary Driver Turbine blade age combined with high flight cycles is a significant predictor of engine failures across the entire fleet Final Statement Environmental factors are a minor, contributing cause but not a primary driver The data provides a clear mandate for action focused on maintenance provider performance and an age based component replacement strategy Slide 9 The Strategic Implication Maria processes the findings So we have A turbine blade crisis in aircraft over 10 years A maintenance quality problem with one provider An interaction effect that multiplies the risk Five aircraft that are essentially flying time bombs Sarah Williams does the math Replacing all old turbine blades $5M Finding a new provider transition cost $2M But if we don't Marcus completes her thought My model predicts 3 2 major failures in the next 6 months if we change nothing At 500K per day in penalties, plus reputation damage Slide 10 Maria Makes the Call Maria's Decision I want three things A detailed inspection protocol for every QuickTurn maintained aircraft A blade replacement schedule based on the regression model A real time monitoring system using Marcus' predictive model The Strategic Proposal We're going to propose a conditional acceptance to Horizon we'll take the contract, but only with a phased approach based on these statistical insights Prediction Accuracy 89 Critical Aircraft 5 Slide 11 Data Driven Decision Making What We Discovered Primary Factors Turbine blade age (28 importance) Maintenance provider quality (22 ) Component interaction effects Statistical Confidence 89 prediction accuracy 71 variance explained (R 2 ) Massive effect sizes confirmed Through rigorous statistical analysis, we transformed uncertainty into actionable intelligence, enabling evidence based decisions that will save millions and protect lives Tutorial The Investigation Deepens Transcript This document provides a complete text based transcript of the interactive tutorial Part 1 Is Blade Age the Primary Failure Driver Introduction The first hypothesis is that older turbine blades fail more often We use a formal hypothesis test to see if the data supports this Section 1 The Hypothesis Test Null Hypothesis (H) Blade age has no effect on failure rate Alternative Hypothesis (H) Blades 10 years old fail more The analysis yields a p value of 0 0003 Let's test this against our significance level (alpha) of 0 05 Interactive Element A button Is the result significant reveals the following conclusion Since the p value (0 0003) is much smaller than our significance level (alpha 0 05), we Reject the Null Hypothesis This means we have strong statistical evidence that older blades do have a higher failure rate Section 2 Regression Analysis Quantifying the Effect The p value told us the effect is real A regression formula tells us how big that effect is Formula Failure Rate 0 082 (0 0134 Blade Age Years) Interactive Element A calculator allows the user to enter a blade age to see the predicted failure rate For example, for a 12 year old blade Predicted Failure Rate 0 2428 or 24 28 Insight The hypothesis test proved age is a factor, and the regression model quantifies it We can now predict the increased risk for any specific blade age Part 2 Is the Maintenance Provider to Blame Introduction The second hypothesis is that one maintenance provider, QuickTurn, is worse than the others We use ANOVA (Analysis of Variance) to compare the mean performance of all providers at once Section 1 ANOVA Results The ANOVA test gives a p value of Section 2 Post Hoc Tests Pinpointing the Problem To find out which provider is different, we use post hoc tests (like Tukey HSD) to compare each pair directly Interactive Element Buttons for each comparison reveal the following results QuickTurn vs Premier p value QuickTurn vs In house p value QuickTurn vs Industry Avg p value Premier vs In house p value 0 05 Conclusion Not Significantly Different Insight The ANOVA test found a problem, and the post hoc tests pinpointed it QuickTurn is statistically worse than Premier, In house, and the industry average This isn't bad luck it's a proven performance gap Test hypotheses about failure drivers and build predictive models to guide the maintenance strategy Use context and information from this module'sCase Study Scenario Statistical Requirements All tests at 0 05 significance level Report confidence intervals for all estimates Include effect sizes, not just p values Validate model assumptions Use the following data sets for your case study assignments Component Failure Analysis Interaction Effects Data Maintenance Quality Testing Test hypotheses about failure drivers and build predictive models to guide the maintenance strategy Use context and information from this module'sCase Study Scenario Statistical Requirements All tests at 0 05 significance level Report confidence intervals for all estimates Include effect sizes, not just p values Validate model assumptions Hypothesis Testing Report (five pages with Excel backup) Download and use theHypothesis Testing Template (XLSX) Download Hypothesis Testing Template (XLSX) Test at least three hypotheses about failure causes Include the null alternative hypotheses, test selection justification, results, and interpretation Use required tests t test, ANOVA, chi square Include effect sizes and practical significance discussion Regression Analysis Simple regression Include blade age vs failure rate Multiple regression Include at least four predictors Check assumptions linearity, normality, and homoscedasticity Interpret coefficients in business terms Predictive Model Development Build a logistic regression for failure prediction Calculate sensitivity, specificity, and accuracy have a risk scoring system for the fleet Validate using a holdout sample Component Interaction Analysis Test for interaction effects between Blade age x Maintenance provider Aircraft age x Cycles Conversion year x Current utilization Visualize interactions with appropriate plots Strategic Recommendations (two page executive summary) Any aircraft that needs immediate intervention ROI of replacing all old blades Perspective on whether AAS should blacklist Quick Turn from the contract Your confidence in any recommendations

Question

CASE STUDY SCENERIO Hypothesis Testing Transcript This document provides a complete text based transcript of the interactive tutorial  Introduction  From Data to Decision Hypothesis testing is the structured process of using data to validate or reject an assumption about the real world  It's how we move from  I think  to  I can prove  and make decisions based on evidence, not just intuition  Core Concepts The Null Hypothesis (H)  This is the default assumption, the status quo  It states there is no effect or no difference  For example,  The new training method has no effect on technician speed   Your goal is to find enough evidence to reject this claim  The Alternative Hypothesis (H)  This is what you are trying to prove  It states there is an effect or a difference  For example,  The new training method does improve technician speed   The p value  This is the most critical output  The p value is the probability of observing your data (or something more extreme) if the null hypothesis were true  A small p value (typically 1  The t test  Comparing Two Groups Description  Use a t test when you want to compare the average (mean) of two, and only two, groups  It tells you if the difference between their means is statistically significant or just due to random chance  Apex Scenario  Training Method Effectiveness Question  Is our new simulation based training better than the traditional method  H  Mean completion time is the same for both groups  H  Mean completion time is different for the two groups  Chart Description  A bar chart compares the  Avg  Completion Time (Hours)  for two groups  Traditional Training  31 hours New Simulation  26 hours The tutorial notes that if the p value for this comparison is low ( 2  ANOVA  Comparing More Than Two Groups Description  ANOVA (Analysis of Variance) is an extension of the t test  Use it when you need to compare the means of three or more groups  It checks if there's a significant difference somewhere among the groups by comparing the variation between groups to the variation within each group  Apex Scenario  Maintenance Provider Effect Question  Do different maintenance providers (QuickTurn, Premier, In house) have different impacts on reliability  H  The mean reliability is the same across all providers  H  At least one provider has a different mean reliability  Chart Description  A bar chart compares the  Mean Time Between Failures (Hours)  for three groups  QuickTurn MRO  3,847 hours In House  6,000 hours Premier MRO  7,200 hours The tutorial notes that a low p value from ANOVA tells you that at least one group is different , but not which one  You then use  post hoc  tests (like Tukey's HSD) to find the specific differences  3  The Chi Square () Test  Analyzing Categories Description  Use this test when working with categorical data (names, types, yes no)  It determines if there is a significant association between two categorical variables by comparing the observed frequencies in your data to the frequencies you would expect if there were no relationship  Apex Scenario  Bay Configuration Impact Question  Is there a relationship between the Maintenance Bay used and the occurrence of Rework  H  Bay assignment and rework are independent  H  Bay assignment and rework are associated  Table Description  A table shows the  Observed Frequencies  of rework  Bay 1 (New)  5 Reworks, 95 No Reworks Bay 2 (Old)  15 Reworks, 85 No Reworks The tutorial notes that the Chi Square test calculates if the observed count of  5  reworks in Bay 1 is significantly lower than what you'd expect by random chance  4  Regression Analysis  Finding Relationships   Predicting Outcomes Description  Regression analysis goes beyond simple comparisons  It models the relationship between a dependent variable (the outcome you want to predict) and one or more independent variables (the drivers)  It's incredibly powerful for forecasting and understanding the  why  behind your data  Apex Scenario  Turbine Blade Failure Question  What drives engine failure  Model  Failure Rate     (Blade Age)   (Cycles) Interpretation  The model tells us exactly how much the failure rate increases (the coefficient ) for each additional year of blade age, while controlling for the effect of cycles  Chart Description  A scatter plot shows  Engine Failure Data  with  Blade Age (Years)  on the x axis and  Failure Rate ( )  on the y axis  The points show a clear upward trend  A  Regression Line  is drawn through the data, visually representing the relationship modeled by the formula  The tutorial notes that the R squared value (e g , 0 71) tells you the percentage of variation in the outcome that is explained by the model's inputs  Apex Aviation  The Investigation Deepens Transcript This document provides a complete text based transcript of the interactive presentation  Slide 1  The Investigation Deepens  We need to know WHY  What's driving these failures     Maria Rodriguez Slide 2  The Task Force Reconvenes of Data The Horizon Task Force reconvenes in AAS' engineering lab  The walls are now covered with scatter plots and Pareto charts  Marcus Thompson stands before a massive display showing component failure data  Maria Rodriguez opens   Last week's analysis was sobering  We have a bimodal fleet with four aircraft in critical condition  Now we need to know WHY  What's driving these failures  Marcus begins   I've spent the week testing three hypotheses   Turbine blade age is the primary failure driver QuickTurn MRO's maintenance quality is substandard Cargo conversion stress accelerates deterioration  Let me show you what the data reveals     Slide 3  The Interrogation of Data Marcus Thompson Tests three core hypotheses to find the root cause of failures  Quote   I tested whether turbine blade age predicts engine failures   James Chen Identifies critical patterns in the data that others might miss  Quote   Look at this clustering     Ahmed Hassan Questions the data, ensuring all variables are considered  Quote   But is it just age, or is it cycles   Lisa Park Focuses on external factors and accountability of suppliers  Quote   Can we prove QuickTurn is actually worse, not just unlucky   Sarah Williams Challenges assumptions to prevent costly misinterpretations of the data  Quote   Maybe QuickTurn gets the worst aircraft   Slide 4  Hypothesis 1  The Turbine Blade Investigation Heading  Hypothesis 1  The Turbine Blade Investigation Question  Is blade age the primary driver of engine failures  Section 1  The Statistical Proof Test Result (t test)  p value Conclusion  The effect of blade age on failures is statistically significant  Regression Analysis  R   0 71 Factoid  R squared (R) is a statistical measure that represents the proportion of the variance for a dependent variable that's explained by an independent variable  An R of 0 71 means that 71  of the variation in engine failures can be explained by blade age and cycles combined  Dialogue   Age is the stronger predictor, but cycles matter too  Together they explain 71  of the variance in failures     Marcus Section 2  Interaction Dialogue   Look at this clustering   every engine failure above 50,000 cycles had blades over 10 years old     James Dialogue   But is it just age, or is it cycles  We need to be sure     Ahmed Slide 5  Hypothesis 2  The Maintenance Provider Effect Heading  Hypothesis 2  The Maintenance Provider Effect Question  Is QuickTurn MRO's maintenance quality substandard  Dialogue   Can we prove QuickTurn is actually worse, not just unlucky     Lisa Section 1  The Verdict Finding  ANOVA results show QuickTurn's maintenance leads to a significantly lower MTBF (p Factoid  ANOVA (Analysis of Variance) is a statistical test used to determine whether there are any statistically significant differences between the means of two or more independent groups  A p value Challenge   Maybe QuickTurn gets the worst aircraft     Sarah Response  Marcus's matched pairs analysis controlled for this, confirming the provider effect was real and not just due to them servicing older planes  Slide 6  Hypothesis 3  Cargo Conversion Stress Heading  Hypothesis 3  Environmental Factors Question  Does cargo conversion accelerate deterioration  Section 1  Initial Analysis H  Cargo conversion timing has no effect on reliability H  Earlier conversions lead to worse reliability MTBF   8234   892 Years Since Conversion R   0 23, p   0 08  The relationship exists but isn't as strong as expected     Section 2  Interaction Model  Watch what happens when I add an interaction term     MTBF   9012   234 Years Since Conversion   156 Aircraft Age   88 Years Since Conversion Aircraft Age R   0 58, p Ahmed Hassan interprets   So older aircraft suffer more from cargo conversion stress  The 2021 conversion on 23 year old aircraft were a perfect storm   Key Insight  The cargo conversion effect is not linear   it compounds with aircraft age, creating exponentially worse reliability in older converted aircraft  Slide 7  The Smoking Gun  Component Interaction Analysis Logistic Regression Model logit(P)    4 23   1 34OldBlades   0 89QuickTurn   0 67HighCycles   2 45(OldBladesQuickTurn) When old blades   QuickTurn maintenance  Failure probability increases 11 6x 95  CI  6 2x to 21 7x James Chen stands up   That's it  QuickTurn isn't properly inspecting turbine blades  They're missing the early warning signs   Slide 8  The Predictive Model Using all significant factors to predict failures  Random Forest Model Results  Overall Accuracy  89  Sensitivity  84  Specificity  91  Hour Prediction  1,000 Variable Importance  Blade age  28  Last maintenance provider  22  Cycles since conversion  18  Interaction terms  15  High Risk Aircraft (Next 1,000 Hours) N411HC   87  failure probability N416HC   84  failure probability N417HC   79  failure probability N404HC   72  failure probability N412HC   71  failure probability  I can now predict with 89  accuracy which aircraft will have major failures in the next 1,000 hours     Marcus Thompson The investigation has successfully identified the primary drivers of unreliability in the Horizon fleet  Primary Driver  Substandard maintenance from QuickTurn MRO is the single largest contributor to poor reliability, independent of aircraft age  Secondary Driver  Turbine blade age combined with high flight cycles is a significant predictor of engine failures across the entire fleet  Final Statement  Environmental factors are a minor, contributing cause but not a primary driver  The data provides a clear mandate for action focused on maintenance provider performance and an age based component replacement strategy  Slide 9  The Strategic Implication Maria processes the findings   So we have  A turbine blade crisis in aircraft over 10 years A maintenance quality problem with one provider An interaction effect that multiplies the risk Five aircraft that are essentially flying time bombs  Sarah Williams does the math  Replacing all old turbine blades  $5M Finding a new provider transition cost  $2M  But if we don't     Marcus completes her thought  My model predicts 3 2 major failures in the next 6 months if we change nothing  At 500K per day in penalties, plus reputation damage    Slide 10  Maria Makes the Call Maria's Decision   I want three things   A detailed inspection protocol for every QuickTurn maintained aircraft A blade replacement schedule based on the regression model A real time monitoring system using Marcus' predictive model The Strategic Proposal   We're going to propose a conditional acceptance to Horizon   we'll take the contract, but only with a phased approach based on these statistical insights   Prediction Accuracy  89  Critical Aircraft  5 Slide 11  Data Driven Decision Making What We Discovered  Primary Factors  Turbine blade age (28  importance) Maintenance provider quality (22 ) Component interaction effects Statistical Confidence  89  prediction accuracy 71  variance explained (R 2 ) Massive effect sizes confirmed Through rigorous statistical analysis, we transformed uncertainty into actionable intelligence, enabling evidence based decisions that will save millions and protect lives  Tutorial  The Investigation Deepens Transcript This document provides a complete text based transcript of the interactive tutorial  Part 1  Is Blade Age the Primary Failure Driver  Introduction  The first hypothesis is that older turbine blades fail more often  We use a formal hypothesis test to see if the data supports this  Section 1  The Hypothesis Test Null Hypothesis (H)  Blade age has no effect on failure rate  Alternative Hypothesis (H)  Blades   10 years old fail more  The analysis yields a p value of 0 0003   Let's test this against our significance level (alpha) of 0 05  Interactive Element  A button  Is the result significant   reveals the following conclusion  Since the p value (0 0003) is much smaller than our significance level (alpha   0 05), we Reject the Null Hypothesis   This means we have strong statistical evidence that older blades do have a higher failure rate  Section 2  Regression Analysis  Quantifying the Effect The p value told us the effect is real  A regression formula tells us how big that effect is  Formula  Failure Rate   0 082   (0 0134   Blade Age Years) Interactive Element  A calculator allows the user to enter a blade age to see the predicted failure rate  For example, for a 12 year old blade  Predicted Failure Rate  0 2428 or 24 28  Insight  The hypothesis test proved age is a factor, and the regression model quantifies it  We can now predict the increased risk for any specific blade age  Part 2  Is the Maintenance Provider to Blame  Introduction  The second hypothesis is that one maintenance provider, QuickTurn, is worse than the others  We use ANOVA (Analysis of Variance) to compare the mean performance of all providers at once  Section 1  ANOVA Results The ANOVA test gives a p value of Section 2  Post Hoc Tests  Pinpointing the Problem To find out which provider is different, we use post hoc tests (like Tukey HSD) to compare each pair directly  Interactive Element  Buttons for each comparison reveal the following results  QuickTurn vs  Premier  p value QuickTurn vs  In house  p value QuickTurn vs  Industry Avg  p value Premier vs  In house  p value  0 05  Conclusion  Not Significantly Different  Insight  The ANOVA test found a problem, and the post hoc tests pinpointed it  QuickTurn is statistically worse than Premier, In house, and the industry average  This isn't bad luck  it's a proven performance gap  Test hypotheses about failure drivers and build predictive models to guide the maintenance strategy  Use context and information from this module'sCase Study Scenario  Statistical Requirements  All tests at   0 05 significance level  Report confidence intervals for all estimates  Include effect sizes, not just p values  Validate model assumptions  Use the following data sets for your case study assignments  Component Failure Analysis Interaction Effects Data Maintenance Quality Testing Test hypotheses about failure drivers and build predictive models to guide the maintenance strategy  Use context and information from this module'sCase Study Scenario  Statistical Requirements  All tests at   0 05 significance level  Report confidence intervals for all estimates  Include effect sizes, not just p values  Validate model assumptions  Hypothesis Testing Report (five pages with Excel backup)  Download and use theHypothesis Testing Template (XLSX) Download Hypothesis Testing Template (XLSX)  Test at least three hypotheses about failure causes  Include the null alternative hypotheses, test selection justification, results, and interpretation  Use required tests  t test, ANOVA, chi square  Include effect sizes and practical significance discussion  Regression Analysis  Simple regression  Include blade age vs  failure rate Multiple regression  Include at least four predictors  Check assumptions  linearity, normality, and homoscedasticity  Interpret coefficients in business terms  Predictive Model Development  Build a logistic regression for failure prediction  Calculate sensitivity, specificity, and accuracy  have a risk scoring system for the fleet  Validate using a holdout sample  Component Interaction Analysis  Test for interaction effects between  Blade age x Maintenance provider Aircraft age x Cycles Conversion year x Current utilization Visualize interactions with appropriate plots  Strategic Recommendations (two page executive summary)  Any aircraft that needs immediate intervention ROI of replacing all old blades Perspective on whether AAS should blacklist Quick Turn from the contract Your confidence in any recommendations

SolutionInn · Accepted Answer

The Answer is in the image, click to view ...

Question: CASE STUDY SCENERIO Hypothesis Testing Transcript This document provides a complete text-based transcript of the interactive tutorial. Introduction: From Data to Decision Hypothesis testing is

Step by Step Solution