Home nlx STEM and Green Task Bundles

STEM and Green Task Bundles

This page compares official occupation labels with postings-derived task-bundle evidence in the NLx corpus. STEM is the main case: the task signal is large enough to support a fuller comparison between official O*NET/BLS STEM occupations and postings-derived STEM candidates. Green is included below as a smaller companion check.

NLx data disclaimer. The National Labor Exchange (NLx) Data Trust bears no responsibility for the analyses or interpretations of the data presented here.
Candidate convention. A postings-derived candidate is an occupation with above-average concentration of the task bundle (RCA > 1) and enough supporting evidence: at least 25,000 active jobs, 250 task mentions, 5 distinct task IDs, 0.08 task mentions per active job, and a 4% task share. Occupations with RCA > 1 that fail one or more of those checks are shown as “not enough evidence for candidate.” These are diagnostic posting measures, not replacements for official classifications.

STEM And Green Task Intensity

LLM-validated task mentions are divided by NLx monthly active jobs. This normalizes by the MAJ denominator rather than raw posting volume.

Compare changes in STEM- and green-task intensity over time, then inspect their distribution across occupations below.

Figure NS-1. STEM and green task intensity over time

Loading...

STEM Task Bundle

STEM task bundle. O*NET STEM task statements appearing in NLx postings, aggregated across SOC-2 groups. The figure is a visual guide to the task bundle used below; it is not a STEM-occupation label.
No-rim chord map of O*NET STEM task bundle links across SOC-2 occupation groups

Official STEM occupations

164

Posting-derived STEM candidates

72

Overlap

47

Official only

117

Posting-only

25

Corpus STEM task share

2.5%

Table 1. Official STEM and postings-derived STEM candidates

Terms. Group compares official O*NET/BLS STEM labels with the posting-derived STEM screen. Active jobs is the NLx active-job denominator; mentions are LLM-validated STEM task mentions.
No Results
STEM candidate screen

STEM_task_RCA = occupation STEM task share / corpus STEM task share

postings_derived_STEM_candidate = 1 when STEM_task_RCA > 1, active_jobs >= 25,000, STEM_task_mentions >= 250, distinct_STEM_task_IDs >= 5, STEM_mentions_per_active_job >= 0.08, and STEM_task_share >= 4%.

Table 2. Top occupations by LLM-validated STEM task RCA

Terms. RCA = occupation STEM task share / corpus STEM task share. Share = STEM task mentions / all task mentions for the occupation. Per job = STEM task mentions / active jobs. “Not enough evidence for candidate” means RCA > 1, but one or more candidate evidence checks failed.
No Results

Green Task Bundle Companion Check

The table below compares official green-occupation labels with the posting-derived green-task measure.

Green task bundle. O*NET green task statements appearing in NLx postings, aggregated across SOC-2 groups. This is a compact companion view for the green task bundle, not a green-occupation label.
No-rim chord map of O*NET green task bundle links across SOC-2 occupation groups

Official green occupations

83

Posting-derived green candidates

9

Overlap

6

Official only

77

Posting-only

3

Corpus green task share

0.2%

Table 3. Top occupations by LLM-validated green task RCA

Terms. RCA, Share, and Per job follow the STEM table definitions, replacing STEM task mentions with LLM-validated green task mentions. “Not enough evidence for candidate” means RCA > 1, but one or more candidate evidence checks failed.
No Results

Evidence Path

STEM and green task labels use separate two-model agreement procedures documented in Agreement And Audit Trail. RCA is defined in Methods.