book

Regression Analysis Microsoft® Excel®

by Conrad Carlberg

May 2016

Beginner to intermediate

368 pages

11h 41m

English

Que

Read now

Unlock full access

How Variation Is MeasuredThe Standard DeviationThe Standard Error of the Mean
Measuring CorrelationCalculating CorrelationCorrelation and CausationRestriction of Range
Predicting with Correlation and Standard ScoresPredicting with Regression Coefficient and InterceptShared VarianceThe TREND() FunctionPartial and Semipartial Correlations
Array-Entering LINEST()Comparing LINEST() to SLOPE() and INTERCEPT()The Standard Error of a Regression CoefficientThe Squared Correlation, R2The Standard Error of EstimateUnderstanding LINEST()’s F-ratioThe General Linear Model, ANOVA, and Regression AnalysisOther Ancillary Statistics from LINEST()
A Composite Predictor VariableUnderstanding the TrendlineMapping LINEST()’s Results to the WorksheetBuilding a Multiple Regression Analysis from the Ground UpUsing the Standard Error of the Regression CoefficientUsing the Models Comparison Approach to Evaluating PredictorsEstimating Shrinkage in R2
About AssumptionsThe Straw ManCoping with Nonlinear and Other Problem DistributionsThe Assumption of Equal SpreadUnequal Variances and Sample Sizes
Dummy CodingEffect CodingOrthogonal CodingFactorial AnalysisStatistical Power, Type I and Type II ErrorsCoping with Unequal Cell Sizes
Contrasting the ResultsStructuring a Conventional ANCOVAStructuring an ANCOVA Using RegressionChecking for a Common Regression LineTesting the Adjusted Means: Planned Orthogonal Coding in ANCOVAANCOVA and Multiple Comparisons Using the Regression ApproachMultiple Comparisons via Planned Nonorthogonal ContrastsMultiple Comparisons with Post Hoc Nonorthogonal Contrasts

Content preview from Regression Analysis Microsoft® Excel®

6. Assumptions and Cautions Regarding Regression Analysis

In This Chapter

About Assumptions

The Straw Man

Coping with Nonlinear and Other Problem Distributions

The Assumption of Equal Spread

Unequal Variances and Sample Sizes

If you bear with me through this entire book, or even just some of it, you’ll find that I refer occasionally to assumptions that your data may need to meet if your analysis is to be credible. Some sources of information about regression analysis—really outstanding sources—spend less than a page discussing those assumptions. Other sources spend pages and pages warning about violating the assumptions without explaining if, how, and why they can be violated with impunity.

The truth of the matter is that assumptions in regression ...