ESM Preprocessing Gallery

Welcome to the R Gallery for preprocessing data from ESM studies!

This website, based on Revol et al. (in preparation), is a comprehensive resource for those interested in preprocessing data from ESM studies. The purpose is to help researchers in navigating the ESM preprocessing steps (ESM preprocessing framework), by providing them with helpful tools and resources (R code and functions) and assisting them in reporting this critical step effectively. Moreover, we provide guidelines and templates for clearly and concisely report the data preprocessing and the characteristics of the preprocessed dataset. Basic knowledge of R, and the tidyverse suite (e.g., dplyr, ggplot2) is required, and some adaptation of the code to your specific situation may be necessary. This website also comes with its own R package, the esmtools R package. Please remember to cite us if you find our framework and resources helpful in your study. Additionnaly, this website and associated package are open to contributions and will undergo regular updates.

Step 1: Import data and preliminary preprocessing

This section is dedicated to the importation of the dataset(s), the first look and the preliminary preprocessing methods needed before or after merging data sources.

Import data

First glimpse

Reformating

Duplication

Renaming and relabeling

Flag (in)valid observations

First missingness analysis

Create time variables

Merging data sources

Check variable coherence

Dataframe format

Branching items

Step 2: Design and Sampling scheme

This section is dedicated to checking if the design and the sampling scheme of the study have been well followed during data collection.

Coherence timestamps and obs order

Calendar plot

Sampling scheme plot

Duration plot

Incoming

Check sampling scheme

Time sent beeps

Delay to send

Quantity of beeps sent

Delay two sent beeps

Dyadic delay in sending

Step 3: Participants Response Behaviors

This section is dedicated to checking whether the participants adopted response behaviors that could decrease data quality and quantity (e.g., response time, careless responding).

Sampling scheme plot

Delay to start and to fill

Time started and missed

Compliance rate

Delay two answered beeps

Response rate

Item response interact

Item response time

Dyadic time interval

Incoming

Missingness correlates

Careless responding

Visualizing and handling

Longstring

Response time

Incoming

Mode consistency

Step 4: Compute and transform variables

This section is dedicated to the computation of the scores of interest and any transformation applied to them (e.g., person-mean centering) for later visualizions or statistical models.

Centering

Check computation

Lag scores

Common scores

Affect scores

Incoming

Systemic/dyadic variables

Incoming

(De)Trend

Step 5: Descriptive statistics and visualization

This section is dedicated to the description and visualization of the variable of interest as well as the investigation of within and between participant differences.