5. How is a job selected for inclusion? What is the sample size?

QUESTION: How is a "job" selected to be included in the Assessor Series®? What is the wage and salary survey sample size? How do we know that a valid sample size was used to calculate the Assessor Series area wage differentials?

ERI has adopted the DOT’s definition of an “occupation,” used for almost 70 years by professionals, courts and experts and also has added a selection/use threshold based on the generally accepted “norms” of US Administrative Law (ALJ) courts. “Job” equals “occupation.” When ERI talks of a specific job with an incumbent, it may use the term “position.”

This differs from O*NET’s “occupational groups” which are job groups, often called “job families” (eDOT also defines “occupational groups” with its first three digits of eDOT’s nine digit code). O*NET states it has “800 occupations” (it misuses the term). The DOL/OES/BLS surveys combine data (roll up O*NET stragglers) into 742 OES/SOC job families. The US Census continues this roll-up into 471 Census Occupations (while adding an occupation, “Logisticians”).

ERI’s Assessor Series and PAQ’s Occupational Assessor (eDOT) report on ~2,500 to ~10,000 specific occupations respectively (varying each Quarter by additions of new jobs and compressions of waning jobs). Each must have:

o	A unique “primary duty” requiring an identifiable skill, if existent.

o	Two to three unique additional task statements (with related skills, abilities and/or knowledge required)

o	An ERI eSemantic low score of job match relevancy as compared to other specific eDOT occupations.

o	An occupation must exist in counts exceeding 2,500 incumbents in the national economy for use in the Assessor Series.

A job must be estimated to exist in 250 or more instances in the national economy to be included in eDOT (explaining why eDOT has more jobs than the combined Assessor Series’ occupations and matching generally acknowledged, but unwritten, US Social Security Administration ALJ national guidelines). The estimation of a job’s existence is based on data collected by:

o	They are reported in published salary surveys with listed survey participant names and incumbent counts.

o	Their job titles and/or specific descriptions are found on disability claim forms (California).

o	They have been collected by ERI’s eSpideri’s datamining of specific job postings (SalariesReview’s Find Jobs) and are complete, specific and matched to a bona fide employing entity (matched to ERI’s Potential Employer list); vague, general descriptions from “Company Unknown” are not analyzed.

o	PAQ field job analyst input, including cybernetic use of eDOT, report the existence of these occupations and their work measures and duties as found in an on-site job analysis.

o	Users of SalaryExpert’s PRO+ (which lists over 100,000 job titles), including the US Census Lay Titles. Visitor interest in job titles is noted along with their IP source.

o	Users of SalaryExpert’s salary calculator (almost ½ million searches each month); this is the second most popular of the free data sites (and reports conservative Government job family salary data).

o	Requests for “new titles” whenever the eDOT PC program is first used (in a “submit process”) and requests by ERI and PAQ subscribers.

ERI's goal is to provide its findings of the numbers of incumbents actually engaged in any specific job's work. The eDOT Skills Project creates an estimate of these jobs' existence in any of over 12 million known US employers. The Relocation Assessor's Labor Cost module allows the pricing of these jobs in a plant relocation scenario. Because ERI now collects data from various sources (the five levels described), care is taken not to overstate the numbers of jobs represented (salary surveys may report on the same job, differing collection techniques may retrieve duplicate data, etc.)

When it comes to the Executive Compensation Assessor & Survey and the Nonprofit Compensation Assessor & Tax-Exempt Survey where all SEC 10-K, 8-K, proxies and all IRS Form 990 EOs/PFs/ EZs are read, the Assessor Series is more a 100% census than it is a survey (with a random sample).

For example, in health care, the survey sample size represents all known hospital executive pay levels (as reported on proxies for the for-profit hospitals and all nonprofit hospitals -- missing only those that purposely do not report data as required or where their required federal reports are missing). All data collected for the executive and Salary Assessors are utilized in the Geographic Assessor area wage differential analyses (the national data forms the "x-axis.")

The exact methodology of ERI statistical calculations is proprietary, and the exact combination and weighting varies by geographic area (according to the availability of wage/salary data). That said, what has occurred with area wage differentials is the emergence of differing geographic differentials by job function and industry. Health care geographic differentials, for example, differ remarkably from general private industry differentials. (The Salary Assessor and Nonprofit Comparables Assessor may use differing geographic differentials.) Areas have been traditionally described in the Assessor Series based on a city, county, state or province, and country methodology. Because of ERI's Job Availability Survey (used to identify jobs to be included in the Salary and other Assessors and used in the Occupational Assessor (eDOT) disability determination applications), ERI has begun to use even more finite definitions (in the US) using the US Census Bureau’s TIGER® System.

Addresses of employers, disability claimants or job searchers can be pinpointed using US geographic/address data and matched to the closest potential employers. For the job searcher or outplacement specialist, employers within xx miles of an individual’s residence that are expected to employ a specific job can be identified. For disability determinations, employers with known jobs of certain physical and mental capacities within xx miles can be identified. Labor markets can be defined (and easily described via maps provided by sites such as Yahoo or Google). Competitors can be identified; probable staffing patterns can be estimated. In the US, this is accomplished by using the publicly available databases of the TIGER® databases. (Any address can be assigned a longitude and latitude and, with the known distance to the center of the earth, mileage between points is easily found after the calculation of the angle created at the earth’s core.) Both PAQ’s eDOT and ERI’s Assessor Series allow for searches within a known radius of a specified office or residence or employer's address.