Comparing performance between log-binomial and robust Poisson regression models for estimating risk ratios under model misspecification.

Wansu Chen, Lei Qian, Jiaxiao Shi, Meredith Franklin

June 2018 BMC Med Res Methodol

Synopsis of Social media discussions

Discussions often reference the comparison of the two modeling approaches, with phrases like 'robust Poisson remains unbiased,' and mention practical scenarios such as data truncation and model bias, reflecting a moderate level of technical engagement and interest in the article's methodological insights.

A
Agreement
 Moderate agreement

Most discussions acknowledge the validity of the study, recognizing the robustness of the Poisson model under misspecification.

I
Interest
 Moderate level of interest

Participants show moderate curiosity, often referencing the comparison of regression models and their practical implications.

E
Engagement
 Moderate level of engagement

Comments involve some technical details, such as mentions of bias, truncation, and model assumptions, indicating a moderate level of engagement.

I
Impact
 Neutral impact

The overall tone suggests awareness of the study's relevance but without strong claims about transformative impact on the field.

Social Mentions

YouTube

1 Videos

Twitter

6 Posts

Blogs

2 Articles

News

2 Articles

Metrics

Video Views

119

Total Likes

21

Extended Reach

22,101

Social Features

11

Timeline: Posts about article

Top Social Media Posts

Posts referencing the article

Comparison of Log-Binomial and Robust Poisson Regression for Risk Ratio Estimation

Comparison of Log-Binomial and Robust Poisson Regression for Risk Ratio Estimation

This study compares the performance of log-binomial and robust Poisson regression models in estimating risk ratios, especially when model assumptions are misspecified. The robust Poisson model remains unbiased under various conditions, making it a reliable choice.

Medicine RTCL TV

August 3, 2023

119 views

  • KONDO Masahiro
    @mkondo1042 (Twitter)

    RR推定時にモデルの誤特定下での修正ポアソン回帰と対数二項回帰の比較。収束の問題もあるし、基本的には修正ポアソンでよいと思っているんだけど、どうなんだろう。 https://t.co/6lZfGjoPBf
    view full post

    September 4, 2022

    7

  • pacoramon
    @pacoramon (Twitter)

    RT @kaz_yos: Comparing performance between log-binomial and robust Poisson regression models for estimating risk ratios under model misspec…
    view full post

    February 9, 2020

    4

  • Sato Shuntaro｜佐藤俊太朗
    @Shuntarooo3 (Twitter)

    RT @kaz_yos: Comparing performance between log-binomial and robust Poisson regression models for estimating risk ratios under model misspec…
    view full post

    February 9, 2020

    4

  • yasukunimemo
    @yasukunimemo (Twitter)

    RT @kaz_yos: Comparing performance between log-binomial and robust Poisson regression models for estimating risk ratios under model misspec…
    view full post

    February 9, 2020

    4

  • WillPer
    @wilpertwitt (Twitter)

    RT @kaz_yos: Comparing performance between log-binomial and robust Poisson regression models for estimating risk ratios under model misspec…
    view full post

    February 8, 2020

    4

  • Kazuki Yoshida
    @kaz_yos (Twitter)

    Comparing performance between log-binomial and robust Poisson regression models for estimating risk ratios under model misspecification | BMC Medical Research Methodology | Full Text https://t.co/mdQBEvdtrY
    view full post

    February 8, 2020

    13

    4

Abstract Synopsis

  • The study compares the performance of log-binomial and robust Poisson regression models in estimating risk ratios, especially when the model assumptions are not fully correct (misspecified).
  • Results show that when the link function is misspecified or responses are truncated, the log-binomial model produces biased results, especially with more truncated data and lower response rates, while the robust Poisson model remains unbiased.
  • Overall, under conditions of model misspecification, the robust Poisson regression model is generally more reliable for estimating risk ratios.