Timothy Lash Seminar, February 11, 2026

Timothy Lash Seminar, February 11, 2026

by Timothy L. Lash

Life & Natural Systems Medicine & Health MBNPublic health and preventive medicineMBNSEpidemiology and Medical statistics
🎙 Timothy L. Lash 👥 2K 📅 February 23, 2026 ⏱ 55 min 👁 375 📄 expert opinion 🧭 2026-08-15
Available in: English (current) Français

Keywords

bias analysismisclassificationsensitivityspecificitycase-control study

Summary

In this seminar, Dr. Timothy Lash discusses quantitative bias analysis (QBA) in epidemiology, focusing on exposure misclassification. He emphasizes that while random error is routinely quantified, systematic errors are often ignored, despite potentially dominating uncertainty. He illustrates the ‘good’ of QBA with a simple selection bias example, showing how bias parameters can be used to adjust estimates. The ‘bad’ is that bias parameters are never known with certainty, as demonstrated through a case-control study on antidepressant use and breast cancer risk. The study reported a null association (OR=1.2), but a validation substudy revealed sensitivities and specificities of exposure classification. Using these validation data in a probabilistic bias analysis, the bias-adjusted estimate became 1.7 with a wide interval, suggesting a potential association. The ‘ugly’ arises when bias analysis is misapplied, as the original authors combined sensitivity and specificity values from different validation subgroups, leading to a misleading null result. Dr. Lash concludes that QBA is essential for identifying uncertainty and guiding better study designs, and he emphasizes that later prospective cohort studies have shown no association.

175 words

Critical Evaluation

Value of the Information & Strength of the Argument

The seminar provides high-value information by demonstrating the practical application of quantitative bias analysis with a real-world example. The argumentation is solid, systematically building from basic concepts to a complex case study. Dr. Lash clearly explains the mathematical underpinnings and the importance of specificity in misclassification bias. He critically evaluates the original study’s bias analysis, highlighting potential misuse and the impact of assumptions. The reasoning is transparent and well-supported by the presented data.

Scientific Rigor, Source Quality, Title Accuracy

The seminar is scientifically rigorous, referencing specific studies (e.g., Boudreau et al., Chan et al.) and methodological literature. The speaker is a recognized expert, and the content aligns with established epidemiological methods. The title accurately reflects the content. The description provides a link to the department’s degree program, which is relevant but not directly a source for the seminar’s claims.

148 words

Title / Content Match

The title accurately reflects the seminar content, which focuses on quantitative bias analysis in epidemiology.

Quality & Reliability

9/10

The speaker is a distinguished professor of epidemiology with extensive expertise in quantitative bias analysis. The seminar is based on peer-reviewed literature and demonstrates rigorous methodological reasoning. The content is well-structured and transparent about uncertainties.

Key Moments

Markers derived by PSI from the transcript: the creator did not define chapters.

Cited Sources

Concurring Sources

Dissenting Sources

  • Original study on antidepressant use and breast cancer risk — The original study's bias-adjusted estimate (1.2) differs from the re-analysis (1.7) due to different assumptions about sensitivity and specificity.

Contribution & Novelties

The seminar provides a clear and practical demonstration of quantitative bias analysis, emphasizing the importance of considering systematic errors in epidemiological studies. It highlights the often-overlooked role of specificity in misclassification bias and illustrates how validation data can be used to adjust estimates. The case study serves as a cautionary tale about the potential for misuse of bias analysis, emphasizing the need for transparency and adherence to methodological principles.

Pour aller plus loin :

  • Quantitative bias analysis — Overview of the method and its applications.
  • Misclassification bias — Explanation of misclassification and its impact on study results.
  • Bias analysis in epidemiology — A review of bias analysis methods and their use in epidemiology.

113 words

Radar Profile

The radar profile shows high scores in information quantity, quality, and technical level, with slightly lower but still high reliability. This indicates a technically advanced and reliable seminar, suitable for an expert audience.

Reliability 9/10

💬 No comments were provided for analysis.