MktPraxis
Interactive Textbook Module Syllabus

Data Preparation and Basic Analysis

From raw survey data to findings: data cleaning (speeders, bots, straightliners), coding, missing data, descriptive statistics, cross-tabulation, hypothesis testing, correlation, regression, and AI-powered analysis — with a full case application.

⏱️ ~44 min reading & interactives 📑 8 sections 🎓 Level: Core undergraduate 📐 Math: Interpret output 📚 Courses: Marketing Research Methods, Marketing Analytics 🏷️ Topics: Data preparation & cleaning, Descriptive statistics
Full Interactive Chapter (Account Required) The complete chapter reading, embedded live mini-widgets, inline Check-Your-Understanding questions, and progress tracking are available to enrolled students and verified instructors.
Open in Textbook Reader (Sign In) Request Free Instructor Preview

Chapter Section Outline (8 Sections)

Topics and managerial frameworks covered in Data Preparation and Basic Analysis:

  1. 1. The Unglamorous Foundation — Why Data Prep Matters More Than Analysis
  2. 2. Cleaning and Editing — The First Pass
  3. 3. Coding and Variable Creation
  4. 4. Descriptive Statistics — What You Should Always Report First
  5. 5. Hypothesis Testing — The Logic of Statistical Inference
  6. 6. Correlation, Regression, and the Causal Inference Trap
  7. 7. AI-Powered Analysis — What Changed and What Didn't
  8. 8. Case Application — FreshBlend Smoothie Subscription

Interactive Concept Mini-Widgets Embedded in This Chapter (3)

Students manipulate these interactive visual simulations directly inside the chapter reading before answering gated check-in questions:

Exclusion Decision Lab

~100 respondents shown as a dot grid. Some carry quality flags (speeder, straightliner, attention-fail), and a subset of flagged dots are actually VALID respondents.

Teaches: a flag is a suspect, not a verdict; over-aggressive cleaning biases data as badly as keeping bots; converging evidence + pre-specified rules + a sensitivity check keep you honest.

Data Cleaning Detective

Show a respondent's data profile. Student decides: keep, flag, or remove — and why.

Teaches: No single quality flag justifies deleting a respondent; remove only on converging evidence (speed, straightlining, failed checks, junk open-ends) and flag the ambiguous cases instead.

Median-Split Power Drain

A scatterplot of a continuous predictor (income) vs. outcome (spend) with a genuine relationship.

Teaches: dichotomizing a continuous variable quietly throws away statistical power — the one-way-door collapse from the measurement-level thread has a concrete, costly price. (For a bivariate normal, splitting one variable at the median multiplies the correlation by ~sqrt(2/pi) ≈ 0.80, so R^2 falls ~36% and you need ~1.57x the sample to recover the same power.)