8274114 Using the soccer occupation auto-coding tool to help participants self-code their occupation in online questionnaires

<h3>Objective</h3> Coding free-text job descriptions is time-consuming and expensive. We evaluated the use of the SOCcer auto-coding tool to help participants self-code their occupations in an online questionnaire of the population-based Connect for Cancer Prevention Study. <h3>Material and Methods</h3> We used SOCcer to provide participants with a list of the four best-fitting occupation categories based on self-reported job title. Participants were asked to select the best category or ‘none of the above’ (v1). In a second version of the questionnaire, we incorporated answers to a work task question and expanded the list to six suggestions (v2). An expert coded 1,000 randomly selected ‘none of the above’ jobs from each version to the 6-digit level where possible, blind to SOCcer’s suggestions. We examined the frequency of selecting a SOCcer-suggested code. We also evaluated the expert’s ability to code the ‘none of the above’ jobs and compared the expert’s assigned code to SOCcer’s suggested list. <h3>Results</h3> V2 modifications reduced the selection of ‘none of the above’ from 17% of 23,699 v1 jobs to 14% of 12,060 v2 jobs. The ‘none of the above’ jobs had lower median SOCcer scores (v1=0.15; v2=0.16) than the participant-coded jobs (v1=0.53; v2=0.56). For these ‘none of the above’ jobs, incorporating task increased the proportion of jobs the expert could code to 6-digits from 68% in v1 to 95% in v2 and decreased un-codable jobs from 16% to 3%; the remainder were assigned a less detailed code. The expert-assigned code was within the SOCcer-suggested list 27% and 54% of the time in v1 and v2, respectively. <h3>Conclusion</h3> Participants successfully self-coded most jobs, with improved performance observed when we added work tasks and expanded the list of suggested occupations. Incorporating task also improved the expert’s ability to assign the ‘none of the above’ jobs to a detailed occupation code.

Authors

Institutions

Publication Details

Published
2025-10-01
DOI
https://doi.org/10.1136/oemed-2025-epicohabstracts.215
Primary Topic
Telemedicine and Telehealth Implementation
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

8274114 Using the soccer occupation auto-coding tool to help participants self-code their occupation in online questionnaires

D. Russ, Jonas S. Almeida, Melissa C. Friesen, Pabitra R Josse et al.
Telemedicine and Telehealth Implementation
article

8274114 Using the soccer occupation auto-coding tool to help participants self-code their occupation in online questionnaires

D. Russ, Jonas S. Almeida, Melissa C. Friesen, Pabitra R Josse, Nicole Gerlanc, Laura E. Beane Freeman, Leila Orszag
article en

Abstract

<h3>Objective</h3> Coding free-text job descriptions is time-consuming and expensive. We evaluated the use of the SOCcer auto-coding tool to help participants self-code their occupations in an online questionnaire of the population-based Connect for Cancer Prevention Study. <h3>Material and Methods</h3> We used SOCcer to provide participants with a list of the four best-fitting occupation categories based on self-reported job title. Participants were asked to select the best category or ‘none of the above’ (v1). In a second version of the questionnaire, we incorporated answers to a work task question and expanded the list to six suggestions (v2). An expert coded 1,000 randomly selected ‘none of the above’ jobs from each version to the 6-digit level where possible, blind to SOCcer’s suggestions. We examined the frequency of selecting a SOCcer-suggested code. We also evaluated the expert’s ability to code the ‘none of the above’ jobs and compared the expert’s assigned code to SOCcer’s suggested list. <h3>Results</h3> V2 modifications reduced the selection of ‘none of the above’ from 17% of 23,699 v1 jobs to 14% of 12,060 v2 jobs. The ‘none of the above’ jobs had lower median SOCcer scores (v1=0.15; v2=0.16) than the participant-coded jobs (v1=0.53; v2=0.56). For these ‘none of the above’ jobs, incorporating task increased the proportion of jobs the expert could code to 6-digits from 68% in v1 to 95% in v2 and decreased un-codable jobs from 16% to 3%; the remainder were assigned a less detailed code. The expert-assigned code was within the SOCcer-suggested list 27% and 54% of the time in v1 and v2, respectively. <h3>Conclusion</h3> Participants successfully self-coded most jobs, with improved performance observed when we added work tasks and expanded the list of suggested occupations. Incorporating task also improved the expert’s ability to assign the ‘none of the above’ jobs to a detailed occupation code.

Department of Health and Human Services (US)
Openalex Percentile: Top 63%
Telemedicine and Telehealth Implementation
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.