Loved by learners at thousands of companies
How can we measure something like “brand loyalty?” It’s an obvious concept of interest to marketers, but we can’t quite take a ruler to it. Instead, we can design and analyze a survey to indirectly measure such a so-called “latent construct.” In this course, you’ll learn how to design and analyze a marketing survey to describe and even predict customers’ behavior based on how they rate items on “a scale of 1 to 5.” You’ll wrangle survey data, conduct exploratory & confirmatory factor analyses, and conduct various survey diagnostics such as checking for reliability and validity.
Preparing to analyze survey dataFree
In this chapter we will explore the use of surveys in marketing research and the importance of reliability and validity in measurement. We will begin the scale development process and perform exploratory data analysis on freshly-collected survey data.Surveys in marketing research50 xpMeasuring expert agreement100 xpInter-rater reliability100 xpContent validity100 xpMeasurement, validity & reliability50 xpVisualizing response frequencies100 xpReverse-coding items100 xpDescribing survey results50 xpMissing values100 xpExploring item correlations100 xpPreparing the brand reputation survey100 xp
Exploratory factor analysis & survey development
Now that we have cleaned and summarized our survey results, we will look for hidden patterns in the data using exploratory factor analysis. These patterns form the basis of developing “factors” of an unobserved or latent variable. Of particular interest in this stage of survey development is internal reliability, or whether similar items in the survey produce similar scores.What is a latent variable?50 xpFrom correlations to factors100 xpBuilding your first EFA100 xpEFA: How many factors?100 xpEFA & item refinement50 xpRefining the brand reputation survey100 xpComparing EFA model fits100 xpEFA model iteration100 xpAssessing internal reliability50 xpMeasuring coefficient (Cronbach's) alpha100 xpCoefficient alpha by dimension100 xpSplit-half reliability100 xpMeasuring loyalty100 xp
Confirmatory factor analysis & construct validation
Does the data as collected agree with prior beliefs about the latent variable of interest? In this chapter, we will use confirmatory factor analysis to formally test the hypothesis that our model fits our data. We can now answer the question of construct validity, or: “are we really measuring what we are claiming to measure?”CFA & EFA50 xpFactor loadings in EFA & CFA100 xpBuilding a CFA in lavaan100 xpA not-so-good CFA100 xpCFA assumptions & interpretation50 xpAdjusting for non-normality100 xpComparing models using absolute fit measures100 xpComparing CFA models using ANOVA100 xpGroup CFA100 xpConstruct validity50 xpConstruct validity & model fit100 xpConstruct validity & reliability100 xpDeeper into AVE & CR100 xpCFA of the brand reputation survey100 xp
Criterion validity & replication
In this chapter we will use our newly validated scale to test its relationship to demographic variables like age or spending habits. We will also confirm the reproducibility and replicability of the survey. Finally, we will explore the power of factor scores in modeling customer behavior.Concurrent validity & model diagrams50 xpPreparing a scaled data frame100 xpPlotting and analyzing a concurrent validity model100 xpConcurrent validity & Likert-style items100 xpPredictive validity & factor scores50 xpStatistical significance & r-square100 xpPrediction & causation100 xpExploring factor scores100 xpFactor scores & regression100 xpRepeated measures, replication & factor scores50 xpTest-retest reliability100 xpCFA, EFA & replication100 xpWrap-up: from generation to replication...50 xp
DatasetsBrand reputation surveyBrand quality surveyCustomer satisfaction surveyCustomer loyalty survey
PrerequisitesIntroduction to Regression in R
Data analytics educator
George works as an independent analyst and data analytics educator with the goal to help clients manage their data so they think more creatively. He holds a master’s degree in information systems with certificate of achievement in quantitative methods from Case Western Reserve University. He is also a technical expert and lead curriculum developer for Thinkful’s data analytics program. George blogs about data, innovation and career development at georgejmount.com.