Technology-mediated interventions to promote physical activity are growing in popularity and appear to be effective for supporting continued adherence for some people.  One potential mechanism of impact is the cultivation of motivation that is self-determined (i.e., autonomous), which is posited to arise from the satisfaction of three basic psychological needs: competence, relatedness, and autonomy. The Peloton Instructor Needs-Supportive Coaching (PINC) tool was used to code 80 Peloton cycling classes across 4 different class types (Beginner, Power Zone, Groove, and High-Intensity Interval Training) to quantify the frequency of needs-supportive and needs-indifferent coaching within a class.  This dataset provides the raw data for the 80 cycling classes and the calculated percentages for aggregated coaching categories.  

Study data were collected and managed using Research Electronic Data Capture (REDCap) tools hosted at the first author’s institution.  REDCap is a secure, web-based software platform designed to support data capture for research studies, providing 1) an intuitive interface for validated data capture; 2) audit trails for tracking data manipulation and export procedures; 3) automated export procedures for seamless data downloads to common statistical packages; and 4) procedures for data integration and interoperability with external sources.  The data were generated using the Peloton Instructor Needs-Supportive Coaching (PINC) tool to code four different cycling class types: Beginner, High-intensity Interval Training (HIIT), Groove, and Power Zone. Beginner rides are specifically targeted to new riders, Power Zone rides coach riders to align their effort within the class to their own fitness level, Groove rides emphasize music and choreography, and HIIT rides are designed for established riders to get a time-efficient and challenging workout.  Detailed information describing the development of the PINC can be found in the associated manuscript.  The PINC was used by five trained coders to document minute-by-minute the presence or absence of an instructor’s coaching statement that met the definition for each of the needs-supportive coaching categories (competence, relatedness, and autonomy) or needs-indifferent coaching categories in cycling classes.  Within each 60-second epoch, coders were instructed to mark a coaching category as present the first time that an instructor delivered a comment that met the coding criterion.  Coders also documented word-for-word the comment that was coded, to facilitate later coding comparisons.  Only one instance of each coaching category per 60-second epoch was coded.  Thus, a single minute of class time might contain multiple categories of coaching, but coders only documented a single instance of each coaching category within that minute. All cycling classes were double-coded. Coders then met virtually in pairs to review their respective codes and identify “missed” codes (i.e., coaching categories overlooked by one coder, which were harmonized in the database) and “unclear” codes (i.e., codes where coders identified different coaching categories for the same comment and actually disagreed). All unclear coding disagreements were then reviewed by the entire research team, which resolved the disagreements by consensus.

The result of the coding process described above was a database documenting the minute-by-minute presence or absence of needs-supportive and needs-indifferent coaching categories across 80 Peloton cycling classes distributed equally across four class types:  Beginner, Power Zone, Groove, and HIIT.  When coding of the 80 classes was complete, the data were exported from REDCap into SPSS format for data analysis. 

To yield variables for analyses, the percent of class minutes during which a class featured each type of coaching was computed.  For example, the number of minutes during which an instructor provided at least one instance of coaching addressing relatedness was summed and divided by the number of minutes in the class to yield the percent of class time that featured coaching relevant to the need for relatedness.  Coded minutes were comprised of either 20- or 30-minute active time plus one minute of warm-up and one minute of cooldown.  This process resulted in variables reflecting the number of minutes and percent of class time devoted to coaching relevant to relatedness (relatedness; perc_relatedness), competence (competence; perc_competence), autonomy (autonomy; perc_auto), and needs-indifferent coaching (indifferent; perc_indiff).  With respect to relatedness, multiple subtypes of coaching counted toward this variable (i.e., shout-outs, statements of connectedness, and personal sharing), so the dataset contains these lower-order variables as well (i.e., shoutout and perc_shout; connection and perc_connect; sharing and perc_share).  With respect to competence, two coaching subtypes counted toward this variable: acknowledgment, perc_acknowledge; and encouragement, perc_encourage.  There were no coaching subtypes for the autonomy variable, but needs-indifferent coaching took several forms, including instructions on speed or resistance (cadence, perc_cadence) and instructions on biomechanics (biomechanics, perc_mechanics).

First look at the README file. The dataset was collected in the Research Electronic Data Capture (REDCap) tool and then exported and analyzed using SPSS version 27. A CSV file for the dataset is provided (Peloton_Database_80_coded_cycling_classes.csv). The data dictionary (Peloton_Database_80_coded_cycling_Classes_DataDictionary.csv) explains the variables found in the dataset. 

For those with SPSS, the SPSS data file is "Peloton_Database_80_coded_cycling_classes.sav" and the SPSS syntax can be found in "SPSS_syntax_Peloton_Database_80_coded_cycling_classes.sps"" or "SPSS_syntax_Peloton_Database_80_coded_cycling_classes.txt".

The Coding Guide used by coders when coding a Peloton class is provided in "Coding_Guide_Peloton_Database_80_coded_cycling_classes.txt" for context.