Identifying Factors Impacting Missingness within Smartphone-based Research: Implications for Intensive Longitudinal Studies of Adolescent Suicidal Thoughts and Behaviors
Analyses of Missing Experience Sampling & Passive Smartphone Sensor Data in a Study of Adolescents at Risk for Suicide
| Sub-directory | Files | Contents |
|---|---|---|
| 0_setup | 1a-d |
Aggregating individual participant files, cleaning of EMA and passive sensor data |
| 0_setup | 2a_compile_daily_dataset.Rmd |
Aggregating and setting up data structures for analysis across modalities (experience sampling, passive sensing, self-report & interview measures). |
| 0_setup | 3_consistency_reliability.Rmd |
Calculations of internal consistency and reliability for self-report measures |
| 1_analysis | 0_descriptive_results.Rmd |
Descriptive analyses, including associations between missingness of |
| 1_analysis | 1a-c |
Aim 1 analyses of time-varying associations between temporal factors (day of week, month, time since baseline) and missing data |
| 1_analysis | 2a-i |
Aim 1 models of between-participant associations between clinical and sociodemographic variables and missing data |
| 1_analysis | 3_supplement_baseline_cox_models.Rmd |
Aim 1 Cox models looking at associations between baseline clinical & sociodemographic variables and time to drop out from smartphone data collection |
| 1_analysis | 4a_clinical_selfreport_predict_missing_over_month.Rmd |
Aim 1 longitudinal analyses of whether changes in self-report clinical measures predict changes in missing data over the next 30 days |
| 1_analysis | 5a_daily_mood_weekly_si_autocor.Rmd |
Models of autocorrelation of daily mood, weekly suicidal ideation frequency, and their missingness |
| 1_analysis | 6a-d |
Aim 2 analyses of proximal passive smartphone sensor predictors of missing daily mood surveys |
| 1_analysis | 7_missing_predict_followup.Rmd |
Aim 3 analyses of associations between missing data during the study period and self-report clinical outcomes at 6-month follow-up |
| 1_analysis | 8_missing_over_month_predict_clinical_selfreport.Rmd |
Aim 3 longitudinal analyses of whether changes in missing data over 30-day periods predict subsequent changes in clinical self-report measures |
| 1_analysis | 9a-b |
Aim 3 analyses of whether changes in missingness precede or follow changes in weekly suicidal ideation frequency or daily mood |
| 1_analysis | 10_suicide_events_during_study.Rmd |
Aim 3 analyses of between-participant associations between missing data and suicidal events during the 6-month study period |
| 1_analysis | 11a_days_since_last_survey.Rmd |
Supplemental analyses of missingness of passive smartphone sensor data as a function of time since the last observed daily mood survey response |
| 1_analysis | 11b_covid_school_closure.Rmd |
Supplemental analyses of missing data as a function of local public high school closures due to Covid-19 |
| 1_analysis | 12a-b |
Predictive modeling analyses of misisng survey data using train/test splits |
| 1_analysis | model_wrapper_functions.R |
Custom helper functions used in modeling, plotting, and wrangling data across multiple analyses |
| 1_analysis | run_all.sh |
Knits all markdowns in the 1_analysis folder |