Guides

Methods and evidence

Causal claims, effect sizes, survey measurement and the use of language models in research, with the evidence behind each.

• Causal claims from observational data in marketing research
  When can archival, platform or survey data support a causal claim? What ad experiments that tested observational methods found, and how to word your results.
• Effect sizes and meta-analysis in marketing research
  How marketing reports effect sizes (elasticities, r, d), how to compare them with published benchmarks, and how a meta-analysis is built and should be read.
• LLM synthetic respondents in marketing research: evidence and limits
  Can large language models stand in for survey or experiment respondents? Where silicon samples match human data, where they fail, and the risk of AI-completed surveys.
• Survey measurement pitfalls in marketing research
  Common ways survey measures mislead in marketing: common method variance, narrow proxies, usage-driven attitudes, stated intentions, ratio indices and fake respondents.