# Measurement and Modeling of Within-Person Variability in Cannabis Protective Behavioral Strategies: A Novel Approach Using Scale Development, Daily Data, and Machine Learning Methods

> **NIH NIH F31** · UNIVERSITY OF WASHINGTON · 2024 · $48,974

## Abstract

PROJECT SUMMARY/ABSTRACT
 Cannabis is the most commonly used controlled substance in the US.1,2 Young adults (YA; ages 18-25)
report highest rates of use, and recent epidemiological surveys show an increase in both proportion of YA
using cannabis and frequency of use among those that use.3,4 As frequent and heavy cannabis use is
associated with a variety of short- and long-term unwanted physical and psychosocial outcomes (e.g., altered
brain development, impaired judgement and memory, poor educational outcomes),5–7 there is a need for
approaches to help individuals reduce use and/or use-related harms. One approach to mitigating substance-
related harm is using protective behavioral strategies (PBS). PBS for cannabis consist of strategies an
individual can use before, during, after, or instead of using to reduce use or consequence.8,9 In retrospective
assessments, frequency of PBS use is associated with lower past 30-day cannabis use and consequences
and mediates the relation between various risk factors and cannabis outcomes,8,10–15 highlighting PBS as a
promising means of reducing cannabis use and harms. However, research on PBS-focused interventions is
mixed9,16,17; this mixed support may be due to gaps in the PBS literature. Specifically, the majority of PBS
research has consisted of cross-sectional, between-person retrospective designs, thus we lack understanding
about which strategies work for whom under what circumstances. Emerging PBS research suggests both
between- and within-person (i.e., daily) variability in whether an individual uses any PBS, and if so, which
strategies they use. This suggests a need for a daily measure of cannabis PBS to increase understanding of
how and when individuals utilize PBS and under what circumstances PBS are or are not effective. To address
these gaps, the proposed F31 will take a novel, multimethod approach incorporating scale development work,
a daily data study design, and machine learning methods. Specific Aims include (1) developing and validating
a daily measure of cannabis PBS; and (2) utilizing a daily data design and machine learning techniques to
develop models predicting PBS efficacy (reductions in use/consequences) for each strategy for a given
individual. To complement these aims, the applicant will receive training in (1) etiology, prevention,
intervention, and harm reduction methods for substance use, with a focus on cannabis PBS; (2) psychometric
development and quantitative methods including multilevel modeling and machine learning; (3) daily data study
design and methodology; and (4) research dissemination, including manuscript/grant writing and conference
presentations. Study findings will have important implications for future PBS intervention research. Specifically,
results can be used to better understand cannabis PBS on a daily level and improve future technology-based
PBS interventions to reduce cannabis-related harms.

## Key facts

- **NIH application ID:** 10794233
- **Project number:** 5F31DA057796-02
- **Recipient organization:** UNIVERSITY OF WASHINGTON
- **Principal Investigator:** Kirstyn Nicole Smith-LeCavalier
- **Activity code:** F31 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2024
- **Award amount:** $48,974
- **Award type:** 5
- **Project period:** 2023-03-16 → 2025-10-15

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10794233

## Citation

> US National Institutes of Health, RePORTER application 10794233, Measurement and Modeling of Within-Person Variability in Cannabis Protective Behavioral Strategies: A Novel Approach Using Scale Development, Daily Data, and Machine Learning Methods (5F31DA057796-02). Retrieved via AI Analytics 2026-05-25 from https://api.ai-analytics.org/grant/nih/10794233. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*