# modeling strategy

**URL:** https://discourse.datamethods.org/c/stat/rms/10.md

[Latest](https://discourse.datamethods.org/latest.md) · [Categories](https://discourse.datamethods.org/categories.md) · [Tags](https://discourse.datamethods.org/tags.md)

---

## [About the modeling strategy category](https://discourse.datamethods.org/t/about-the-modeling-strategy-category/23)

<div class="topic-metadata">

**Author:** [@f2harrell](https://discourse.datamethods.org/u/f2harrell)\
**Replies:** 0\
**Last updated:** [July 13, 2018, 1:14pm UTC](https://discourse.datamethods.org/t/about-the-modeling-strategy-category/23 "2018-07-13T13:14:37Z")

</div>

General model specification issues, nonlinearities, interactions and heterogeneity of treatment effect, avoiding categorization, how to sequence multiple steps (which may involve multiple imputation and data reduction)

---

## [Bayesian Time Machine](https://discourse.datamethods.org/t/bayesian-time-machine/28884)

<div class="topic-metadata">

**Author:** [@llynn](https://discourse.datamethods.org/u/llynn)\
**Replies:** 0\
**Last updated:** [October 10, 2026, 5:08pm UTC](https://discourse.datamethods.org/t/bayesian-time-machine/28884 "2026-10-10T17:08:46Z")

</div>

I have a question for @Stephen about the “Bayesian Time Machine” (BTM). In the excellent pod cast you warned of the potential for changes in the sampled population over the sequential time intervals (for e…

---

## [The threshold for best prediction is not Youden index](https://discourse.datamethods.org/t/the-threshold-for-best-prediction-is-not-youden-index/28881)

<div class="topic-metadata">

**Author:** [@a.indrayan](https://discourse.datamethods.org/u/a.indrayan)\
**Replies:** 5\
**Last updated:** [October 8, 2026, 3:32am UTC](https://discourse.datamethods.org/t/the-threshold-for-best-prediction-is-not-youden-index/28881 "2026-10-08T03:32:09Z")

</div>

The current practice generally is to use the point with the Youden index as the best threshold. Being based on sensitivity and specificity, this is good for discrimination or classification of already known outcomes. But…

---

## [Bayesian predictive projection for variable selection](https://discourse.datamethods.org/t/bayesian-predictive-projection-for-variable-selection/28620)

<div class="topic-metadata">

**Author:** [@arthur\_albuquerque](https://discourse.datamethods.org/u/arthur_albuquerque)\
**Replies:** 18\
**Last updated:** [September 5, 2026, 11:02am UTC](https://discourse.datamethods.org/t/bayesian-predictive-projection-for-variable-selection/28620 "2026-09-05T11:02:29Z")

</div>

There have been recent advances on the field of (Bayesian) Projection Predictive Inference, especially through the projpred R package. The main goal of this method is to find “the smallest submodel which makes predictio…

---

## [RMS Discussions](https://discourse.datamethods.org/t/rms-discussions/3275)

<div class="topic-metadata">

**Author:** [@f2harrell](https://discourse.datamethods.org/u/f2harrell)\
**Replies:** 395\
**Last updated:** [July 7, 2026, 12:53pm UTC](https://discourse.datamethods.org/t/rms-discussions/3275 "2026-07-07T12:53:00Z")

</div>

This is a place for questions and discussions about the R rms package and for archived discussions arising from Frank Harrell’s Regression Modeling Strategies full or short course and for regression modeling topics from …

---

## [Modeling outcome with 95% probability](https://discourse.datamethods.org/t/modeling-outcome-with-95-probability/28797)

<div class="topic-metadata">

**Author:** [@CP3](https://discourse.datamethods.org/u/CP3)\
**Replies:** 3\
**Last updated:** [June 18, 2026, 6:46pm UTC](https://discourse.datamethods.org/t/modeling-outcome-with-95-probability/28797 "2026-06-18T18:46:51Z")

</div>

I have a dataset where the outcome is binary. 95% yes, 5% no. For whatever reason I decided to fit a fully adjusted model with other covariates like race, age, education and other demographics related variables. One vari…

---

## [RMS Ordinal Regression for Continuous Y](https://discourse.datamethods.org/t/rms-ordinal-regression-for-continuous-y/4816)

<div class="topic-metadata">

**Author:** [@f2harrell](https://discourse.datamethods.org/u/f2harrell)\
**Replies:** 81\
**Last updated:** [June 9, 2026, 9:34am UTC](https://discourse.datamethods.org/t/rms-ordinal-regression-for-continuous-y/4816 "2026-06-09T09:34:27Z")

</div>

Regression Modeling Strategies: Ordinal Regression for Continuous Y This is the 15th of several connected topics organized around chapters in Regression Modeling Strategies. The purposes of these topics are to introduce…

---

## [Collider in RCT Subgroup Analysis](https://discourse.datamethods.org/t/collider-in-rct-subgroup-analysis/28689)

<div class="topic-metadata">

**Author:** [@arthur\_albuquerque](https://discourse.datamethods.org/u/arthur_albuquerque)\
**Replies:** 34\
**Last updated:** [June 7, 2026, 10:55am UTC](https://discourse.datamethods.org/t/collider-in-rct-subgroup-analysis/28689 "2026-06-07T10:55:29Z")

</div>

I came across this interesting article: A proposal for capturing interaction and effect modification using DAGs Interaction: they propose that this term should be used specifically when both treatment and “subgroup” ha…

---

## [RMS Additional Resources](https://discourse.datamethods.org/t/rms-additional-resources/4810)

<div class="topic-metadata">

**Author:** [@f2harrell](https://discourse.datamethods.org/u/f2harrell)\
**Replies:** 3\
**Last updated:** [May 19, 2026, 1:58am UTC](https://discourse.datamethods.org/t/rms-additional-resources/4810 "2026-05-19T01:58:22Z")

</div>

Regression Modeling Strategies: Additional Resources Library of methods and applications papers provided to course participants Regression and Other Stories, Gelman, et al, (2021). Statistical Rethinking: A Bayesian Cou…

---

## [Rank-hazard plot interpretation](https://discourse.datamethods.org/t/rank-hazard-plot-interpretation/28724)

<div class="topic-metadata">

**Author:** [@Julia\_zhan](https://discourse.datamethods.org/u/Julia_zhan)\
**Replies:** 4\
**Last updated:** [April 29, 2026, 8:37pm UTC](https://discourse.datamethods.org/t/rank-hazard-plot-interpretation/28724 "2026-04-29T20:37:27Z")

</div>

Hi Prof Harrell, What is your rationale for using a non-equally spaced y-axis in the rank hazard plot, and how do you address the concern that this may introduce visual distortion? Many thanks in advance. The two figu…

---

## [Experimenting with a Bayesian Partial Proportional Odds Model](https://discourse.datamethods.org/t/experimenting-with-a-bayesian-partial-proportional-odds-model/28677)

<div class="topic-metadata">

**Author:** [@arthur\_albuquerque](https://discourse.datamethods.org/u/arthur_albuquerque)\
**Replies:** 4\
**Last updated:** [April 27, 2026, 12:16pm UTC](https://discourse.datamethods.org/t/experimenting-with-a-bayesian-partial-proportional-odds-model/28677 "2026-04-27T12:16:44Z")

</div>

Imagine we have data from a RCT comparing two strategies (experimental = PCT and control = UC). These strategies aid physicians to decide whether a patient needs treatment with antibiotics or not. We then measure the num…

---

## [Risk factor evaluation in a small surgical sample (N=20, Events=6)](https://discourse.datamethods.org/t/risk-factor-evaluation-in-a-small-surgical-sample-n-20-events-6/28714)

<div class="topic-metadata">

**Author:** [@Yosukesugita](https://discourse.datamethods.org/u/Yosukesugita)\
**Replies:** 3\
**Last updated:** [April 20, 2026, 11:25am UTC](https://discourse.datamethods.org/t/risk-factor-evaluation-in-a-small-surgical-sample-n-20-events-6/28714 "2026-04-20T11:25:42Z")

</div>

Dear all I am evaluating a surgical risk factor in a small retrospective study (N=20, 6 events). The potential risk factor is a continuous variable measured during the operation. I am fully aware of the limitations of s…

---

## [Adjustment in joint models](https://discourse.datamethods.org/t/adjustment-in-joint-models/28547)

<div class="topic-metadata">

**Author:** [@Ivo1](https://discourse.datamethods.org/u/Ivo1)\
**Replies:** 3\
**Last updated:** [April 14, 2026, 2:11pm UTC](https://discourse.datamethods.org/t/adjustment-in-joint-models/28547 "2026-04-14T14:11:18Z")

</div>

I am investigating how one blood biomarker (biomarker\_1, measured at hospital discharge) predicts the longitudinal trajectory of another biomarker (biomarker\_2) in a cohort of patients after discharge. Data structure …

---

## [A/B Testing for Information Gain](https://discourse.datamethods.org/t/a-b-testing-for-information-gain/28690)

<div class="topic-metadata">

**Author:** [@Johannes\_Schwenke](https://discourse.datamethods.org/u/Johannes_Schwenke)\
**Replies:** 1\
**Last updated:** [April 2, 2026, 8:55pm UTC](https://discourse.datamethods.org/t/a-b-testing-for-information-gain/28690 "2026-04-02T20:55:30Z")

</div>

I’d appreciate some thoughts / input on the following: At our institution we will change the software for routine collection of quality of life data which also causes changes in the schedule for data collection. I woul…

---

## [Best model reduction approach for Fine-Gray prediction model](https://discourse.datamethods.org/t/best-model-reduction-approach-for-fine-gray-prediction-model/28678)

<div class="topic-metadata">

**Author:** [@rafdoodle1](https://discourse.datamethods.org/u/rafdoodle1)\
**Replies:** 5\
**Last updated:** [March 27, 2026, 11:27am UTC](https://discourse.datamethods.org/t/best-model-reduction-approach-for-fine-gray-prediction-model/28678 "2026-03-27T11:27:38Z")

</div>

Hello there, I am developing a Fine-Gray competing risks prediction model for dementia incidence (with death as the competing event) using population-based survey data linked to administrative health records. My full mo…

---

## [Estimand when treatment delays disease onset](https://discourse.datamethods.org/t/estimand-when-treatment-delays-disease-onset/28654)

<div class="topic-metadata">

**Author:** [@arthur\_albuquerque](https://discourse.datamethods.org/u/arthur_albuquerque)\
**Replies:** 4\
**Last updated:** [March 5, 2026, 1:08pm UTC](https://discourse.datamethods.org/t/estimand-when-treatment-delays-disease-onset/28654 "2026-03-05T13:08:03Z")

</div>

An interesting trial found that a treatment delayed disease onset, but did not reduce overall incidence at end of follow up: What estimand would you target in such cases? I thought about “difference in time unwell” …

---

## [Narratively generative modeling](https://discourse.datamethods.org/t/narratively-generative-modeling/28644)

<div class="topic-metadata">

**Author:** [@isagonas](https://discourse.datamethods.org/u/isagonas)\
**Replies:** 1\
**Last updated:** [February 24, 2026, 1:24pm UTC](https://discourse.datamethods.org/t/narratively-generative-modeling/28644 "2026-02-24T13:24:06Z")

</div>

I recently came across the idea of narratively generative modeling and have been trying to understand it more deeply, especially the emphasis on specifying a full data-generating process and interrogating it through prio…

---

## [Optimism Correction after LASSO in clinical prediction models](https://discourse.datamethods.org/t/optimism-correction-after-lasso-in-clinical-prediction-models/28579)

<div class="topic-metadata">

**Author:** [@arthur\_albuquerque](https://discourse.datamethods.org/u/arthur_albuquerque)\
**Replies:** 18\
**Last updated:** [January 22, 2026, 2:37pm UTC](https://discourse.datamethods.org/t/optimism-correction-after-lasso-in-clinical-prediction-models/28579 "2026-01-22T14:37:23Z")

</div>

I am developing and internally validating a clinical prediction model. I want to combine LASSO and bootstrapping to get coefficients adjusted for both overfitting and optimism. However, I can’t find references doing both…

---

## [CPM performance metrics: Adjustment for repeated measures](https://discourse.datamethods.org/t/cpm-performance-metrics-adjustment-for-repeated-measures/28603)

<div class="topic-metadata">

**Author:** [@arthur\_albuquerque](https://discourse.datamethods.org/u/arthur_albuquerque)\
**Replies:** 2\
**Last updated:** [January 18, 2026, 2:09am UTC](https://discourse.datamethods.org/t/cpm-performance-metrics-adjustment-for-repeated-measures/28603 "2026-01-18T02:09:41Z")

</div>

I want to develop and internal-external validate a diagnostic clinical prediction multinomial model with a dataset with repeated measures. The project consists of diagnosing multiresistant bacteria in patients with suspe…

---

## [Combining splines with LASSO in clinical prediction models](https://discourse.datamethods.org/t/combining-splines-with-lasso-in-clinical-prediction-models/28595)

<div class="topic-metadata">

**Author:** [@arthur\_albuquerque](https://discourse.datamethods.org/u/arthur_albuquerque)\
**Replies:** 9\
**Last updated:** [January 13, 2026, 12:00pm UTC](https://discourse.datamethods.org/t/combining-splines-with-lasso-in-clinical-prediction-models/28595 "2026-01-13T12:00:57Z")

</div>

I want to include restricted cubic splines while developing a clinical prediction model with a multinomial outcome. However, I also want to apply LASSO for shrinkage. I know that one could fit a grouped-LASSO to ensure …

---

## [Bootstrapped variable selection in the presence of competing risk](https://discourse.datamethods.org/t/bootstrapped-variable-selection-in-the-presence-of-competing-risk/28594)

<div class="topic-metadata">

**Author:** [@OttoPat](https://discourse.datamethods.org/u/OttoPat)\
**Replies:** 7\
**Last updated:** [January 12, 2026, 12:40pm UTC](https://discourse.datamethods.org/t/bootstrapped-variable-selection-in-the-presence-of-competing-risk/28594 "2026-01-12T12:40:22Z")

</div>

Dear DataMethods Community, As a preface: I’m a 1st year medical PhD student and have limited statistical support in my department so forgive me for any potential caveats in my questions I am currently working on devel…

---

## [RMS Missing Data](https://discourse.datamethods.org/t/rms-missing-data/4781)

<div class="topic-metadata">

**Author:** [@f2harrell](https://discourse.datamethods.org/u/f2harrell)\
**Replies:** 42\
**Last updated:** [January 10, 2026, 12:56pm UTC](https://discourse.datamethods.org/t/rms-missing-data/4781 "2026-01-10T12:56:42Z")

</div>

Regression Modeling Strategies: Missing Data This is the third of several connected topics organized around chapters in Regression Modeling Strategies. The purposes of these topics are to introduce key concepts in the c…

---

## [Prediction Recruitment Duration](https://discourse.datamethods.org/t/prediction-recruitment-duration/28583)

<div class="topic-metadata">

**Author:** [@Johannes\_Schwenke](https://discourse.datamethods.org/u/Johannes_Schwenke)\
**Replies:** 2\
**Last updated:** [January 6, 2026, 8:32am UTC](https://discourse.datamethods.org/t/prediction-recruitment-duration/28583 "2026-01-06T08:32:26Z")

</div>

I’m newly involved in a project, where we have access to individual patient recruitment data of 300 trials (\>100k patients) + info on these trials. As most will know, most trials struggle with recruitment so there’s a lo…

---

## [Optimising models for calibration](https://discourse.datamethods.org/t/optimising-models-for-calibration/28405)

<div class="topic-metadata">

**Author:** [@kiwiskiNZ](https://discourse.datamethods.org/u/kiwiskiNZ)\
**Replies:** 51\
**Last updated:** [December 5, 2025, 2:02pm UTC](https://discourse.datamethods.org/t/optimising-models-for-calibration/28405 "2025-12-05T14:02:36Z")

</div>

Whether a regression model or a machine learning model there are times when the performance of the model across certain ranges of probabilities is very important to a clinician making decisions. A lot of what I am invol…

---

## [Feedback on BayeScores: A Bayesian AFT Mixture Cure Model for Quantifying Clinical Benefit](https://discourse.datamethods.org/t/feedback-on-bayescores-a-bayesian-aft-mixture-cure-model-for-quantifying-clinical-benefit/28526)

<div class="topic-metadata">

**Author:** [@albertoca](https://discourse.datamethods.org/u/albertoca)\
**Replies:** 3\
**Last updated:** [November 11, 2025, 1:12pm UTC](https://discourse.datamethods.org/t/feedback-on-bayescores-a-bayesian-aft-mixture-cure-model-for-quantifying-clinical-benefit/28526 "2025-11-11T13:12:30Z")

</div>

Hello datamethods community, My colleague, @Pavlos\_Msaouel and I have been developing a general framework for quantifying the magnitude of clinical benefit in RCTs, aiming to move beyond some of the limitations of hazar…

---

## [Critiquing risk factor analysis](https://discourse.datamethods.org/t/critiquing-risk-factor-analysis/27458)

<div class="topic-metadata">

**Author:** [@trumanfrancis](https://discourse.datamethods.org/u/trumanfrancis)\
**Replies:** 25\
**Last updated:** [October 29, 2025, 11:34am UTC](https://discourse.datamethods.org/t/critiquing-risk-factor-analysis/27458 "2025-10-29T11:34:37Z")

</div>

This is the methodology used to analyze risk factors on a large datset in veterinary medicine. The authors are well published epidemiologists. If you critique the analysis you are met with disdain. What is the best w…

---

## [RMS Multivariable Modeling Strategies](https://discourse.datamethods.org/t/rms-multivariable-modeling-strategies/4782)

<div class="topic-metadata">

**Author:** [@f2harrell](https://discourse.datamethods.org/u/f2harrell)\
**Replies:** 76\
**Last updated:** [October 28, 2025, 3:39pm UTC](https://discourse.datamethods.org/t/rms-multivariable-modeling-strategies/4782 "2025-10-28T15:39:30Z")

</div>

Regression Modeling Strategies: Multivariable Modeling Strategies This is the fourth of several connected topics organized around chapters in Regression Modeling Strategies. The purposes of these topics are to introduce…

---

## [How should I define my exposure variable for a proportional odds logistic regression](https://discourse.datamethods.org/t/how-should-i-define-my-exposure-variable-for-a-proportional-odds-logistic-regression/28375)

<div class="topic-metadata">

**Author:** [@koray\_durak](https://discourse.datamethods.org/u/koray_durak)\
**Replies:** 7\
**Last updated:** [September 21, 2025, 1:55pm UTC](https://discourse.datamethods.org/t/how-should-i-define-my-exposure-variable-for-a-proportional-odds-logistic-regression/28375 "2025-09-21T13:55:05Z")

</div>

Dear all, I am analyzing a surgical cohort where about 10% of patients were on SGLT2 inhibitors before surgery. Among these, roughly half continued therapy perioperatively and half had it paused. The remaining 90% of pa…

---

## [RMS Describing, Resampling, Validating, and Simplifying the Model](https://discourse.datamethods.org/t/rms-describing-resampling-validating-and-simplifying-the-model/4790)

<div class="topic-metadata">

**Author:** [@f2harrell](https://discourse.datamethods.org/u/f2harrell)\
**Replies:** 181\
**Last updated:** [September 13, 2025, 12:01pm UTC](https://discourse.datamethods.org/t/rms-describing-resampling-validating-and-simplifying-the-model/4790 "2025-09-13T12:01:55Z")

</div>

Regression Modeling Strategies: Describing, Resampling, Validating, and Simplifying the Model This is the fifth of several connected topics organized around chapters in Regression Modeling Strategies. The purposes of th…

---

## [Why competing-events are mentioned only in time-to-event framework?](https://discourse.datamethods.org/t/why-competing-events-are-mentioned-only-in-time-to-event-framework/28352)

<div class="topic-metadata">

**Author:** [@Uriah](https://discourse.datamethods.org/u/Uriah)\
**Replies:** 1\
**Last updated:** [August 24, 2025, 8:56pm UTC](https://discourse.datamethods.org/t/why-competing-events-are-mentioned-only-in-time-to-event-framework/28352 "2025-08-24T20:56:24Z")

</div>

I’m curious if anyone else has thought about this issue: Competing events are related to the time-to-event framework, but I don’t recall ever seeing them mentioned as a problem in the binary setting for prognosis predic…

[Next page](https://discourse.datamethods.org/c/stat/rms/10.md?page=1)