Salary Report Research Methodology

Competitive Position® Salary Report

Reproducible

Accepted Standards

The methodology can be followed to reproduce statistically equivalent results.

Accepted statistical standards are applied.

Data Source

Want Ads

The want ad sources are fully documented:

  • names of web site sources
  • beginning and end dates of research
  • number of want ads collected.
Publicly Available

Public availability is one advantage of collecting want ads from career web sites.

Open a browser, find a career web site and read the want ads to become familiar with the source of the data and know precisely what is being analyzed.

Market positions are directly observed as they occur.

Anonymously Obtained

The want ads are anonymously obtained from career web sites.

None of the collected want ads are written for a Salary Report.

Data Extracted

Each want ad collected includes:

  • Salary
  • Experience
  • Qualifications for an IT Job
  • Location in a US City.

Some want ads list a range of salaries and/or experience requirements:

in a want ad with a salary range and one experience requirement -
only the lowest salary is recorded
in a want ad with an experience range and one salary offer -
only the highest experience is recorded
in a want ad with both a salary and an experience range, two separate entries are recorded -
the low salary is recorded with the low experience
the high salary is recorded with the high experience.

Individual want ads are neither identified nor reproduced in the Salary Report.

The Job Market

Good Information

Classified want ads collectively reflect the competitive demand and supply conditions of the job market. Want ads are an unbiased source of salary information.

Employer Demand

Each want ad states an employer's initial bargaining position: potential salary versus desired qualifications. The goal of the listing is to elicit the best qualified candidate within the employer's cost constraints. This goal creates a competitive market for information technology professionals that is displayed within the collection of classified want ads.

The success of a want ad depends on its competitive position relative to other want ads. If the salary is too low relative to the experience requirement, few suitable professionals will submit résumés. If the salary is too high relative to the experience requirement, the employer may not be able to meet the expectations of job candidates without overextending expenses for the desired job performance.

Employers consider their cost-to-performance constraints relative to the salary offers and experience requirements listed in other employer's want ads. The competition between employers for the best résumés restricts the distribution of salary offers relative to experience requirements around central tendency, or average values.

The income level of employers determines their cost constraints and performance needs and, in turn, the demand for information technology professionals. When sales are good for employers the demand for IT professionals is strong and salary offers are high. When sales are bad for employers the demand for IT professionals is weak and salary offers are low. Importantly, the difference in income between employers accounts for some of the variation in salary offers.

Employee Supply

The supply of job candidates determines the level of salary offers needed to entice the best qualified applicants. An excess supply of information technology professionals in the job market is reflected in low salary offers. A short supply forces employers to make high salary offers. Employers adjust their salary offers and experience requirements to reflect the quality of résumés elicited by their want ads.

Competitive Position

The competitive demand for and supply of information technology professionals is reflected in the average, and distribution, of the salary offers and experience requirements of job listings. The average salary offer is determined by the performance needs, sales revenue and cost constraint of employers, in combination with the availability and alternative opportunities of IT professionals. The distribution of salary offers is determined by: the competitive rivalry between employers, the variation in profitability of employers, differences in the quality of skills employers desire, and the location of the job.

Career web sites are today's job market news. Each want ad is written in consideration of an employer's current needs and capabilities relative to their access to qualified job candidates. Collectively, the want ads reflect up-to-date demand and supply conditions.

IT Job

Based on IT Skills

An IT Job title of a Salary Report refers to a set of specific skills in the use of information technology systems and software, such as:

  • programming languages
  • databases
  • applications
  • operating systems
  • networks
  • communications.

For example, "C++ and Java on UNIX - Developer" specifies two programming languages and an operating system.

Developed Empirically

The IT Job titles are developed empirically, from the want ad data itself. An IT Job title is formed when a set of qualifications is required in a significant number of want ads.

The IT Job titles are an indication of the information technology skills in demand. Due to differences in demand, some of the IT Job titles have more specific qualifications than others.

Many Keywords Tracked

Several thousand different keywords are tracked. Information technology terms have been continuously added over the years to form as complete a list as possible. Industry and education requirements are also recorded.

Qualifications Documented

The IT Job qualifications and their associated keywords are listed in each Salary Report.

The qualifications and keywords of additional requirements often found in want ads of an IT Job are also listed in the Salary Report.

US City

US Metropolitan Areas

The Metropolitan Areas are defined by the U.S. Office of Management and Budget (OMB).

The Salary Report can be analyzed in association with other statistical reports that conform with these metropolitan statistical areas. For instance, cost of living data published in the Statistical Abstract of the United States.

Experience

Any Years and/or Months

Any Number of Years and/or Months of Required Experience can be specified -for each IT Job title and US City choice- and a unique Salary Report will be selected. Salary results are available starting at Entry Level.

Statistics

Regression Analysis

Regression analysis is the principal statistical technique of the Salary Report.

The 'best fit' equation is derived, expressing salary as a function of job characteristics. This salary equation possesses the minimum variance between the want ads of the sample.

The job characteristics may include:

  • Experience
  • Date
  • Qualifications required in addition in only some of the want ads of an IT Job
  • US City of some of the want ads of an IT Job
  • IT Job of some of the want ads within a US City
  • IT Job 'type': Application Developer, Lead Application Developer, Systems Professional, Lead Systems Professional, Database Professional, Business Systems Professional or Information Technology Director.
Regression Equation

Two types of equations are derived:

  • IT Job equations may express Salary across the USA as a function of experience, date, extra qualifications and specific US City
  • US City equations may express Salary across information technology jobs as a function of experience, date, IT Job 'type' and specific IT Job

Typically salary is lowest at entry level, increases rapidly with the first years of experience and approaches a ceiling as experience matures - a log-linear relationship between Salary and Experience.

Not all job characteristics have an influence on salary:

  • some characteristics are such an integral part of the job that while all want ads implicitly assume it only some explicitly state it
  • other characteristics may describe the peculiarities of a working environment rather than an especially valued qualification
  • employer's may not have reached a market consensus on the value of a qualification.

The regression equation is fully documented in the Salary Report.

Collinearity Tested

A 95% collinearity test verifies the independent job characteristics.

Only the independent job characteristics that directly influence salaries are specified in the regression equation.

For example, in some want ads a 'senior developer' might refer to a particular position and salary track in the hierarchy of a firm while in other want ads it might only mean more experience. The collinearity test determines whether a 'senior developer' is a separate position or if the years of experience are always greater for want ads requiring it. If 'senior developer' is related to years of experience, then it provides no more information about salary offers beyond what is already known from the years of experience.

F-Distribution Test

The F-Distribution statistical test verifies the significance of the entire regression equation.

The F-Distribution statistic determines the probability that the regression equation adds nothing to the determination of salary. The lower the probability, the more confidence is placed in the regression equation.

The F-Distribution statistical test result are set to a high level of quality for the regression equation. There is usually less than a one-hundredth of a percent probability (<0.01%) that there is no salary offer regression equation.

The F-Distribution statistical test results are fully documented in the Salary Report.

Heteroscedasticity Test

Many want ads state a salary that is either greater or less than the Salary Average. A residual is equal to the difference between the salary offered in a want ad and the salary for a want ad as calculated by the regression equation. The variance is the sum of the squared residuals for the entire sample of want ads.

The variance of the regression equation is tested against Heteroscedasticity.

Heteroscedasticity is a condition where a statistically significant regression equation can be constructed between the variance of the salary offers and the job characteristics in the sample of want ads.

Heteroscedasticity exists when at least one job characteristic influences the residuals. This occurs because employers are less in market consensus on the value of some job characteristics and more in market consensus on the value of other job characteristics.

The difference between high and low salaries across the sample of want ads may be larger:

  • with higher experience requirements
  • from one month to another
  • with an extra qualification
  • in a particular US City
  • in one IT Job 'type'
  • in a specific IT Job.

If employers are less in agreement on the value of particular job characteristics, then this information needs to be incorporated into the salary regression equation.

A regression equation is derived between the variance in salary offers and the job characteristics.

If this heteroscedasticity regression equation is statistically significant, the variance is calculated for each want ad and is applied in a new derivation of the salary offer regression equation. With this heteroscedasticity correction, the 'best fit' salary offer regression equation is found.

If a regression equation can not be derived between the variance in salary offers and the job characteristics, then the original salary offer regression equation is the 'best fit'.

Information about the heteroscedasticity regression equation is documented in the Salary Report.

t-Distribution Tests

The t-Distribution statistical test verifies the significance of each of the components of the regression equation.

The t-Distribution statistic determines the probability that a component contributes nothing to the calculation of salary. The lower the probability, the more confidence is placed on the component in the salary equation.

The t-Distribution statistic is set to a high level of quality for each component in the salary equation. No component is accepted above a t-Distribution probability of 5%. Most components have a t-Distribution probability significantly less than 1%.

The t-Distribution statistical test results are fully documented in the Salary Report.

Geographic Adjustment

US City ÷ USA

The Salary Report provides information about an IT Job in a US City. A geographic adjustment ratio is applied to the IT Job salary numbers.

The geographic adjustment ratio is equal to the US City regression equation divided by the regression equation for the United States of America. Each regression equation is calculated at the required experience of the Salary Report.

The F-Distribution, heteroscedasticity and t-Distribution statistical test results for both the US City and USA regression equations, as well as, the calculation of the geographic adjustment ratio are documented in the Salary Report.

Salary Information

IT Job and US City

A Salary Report presents salary data calculated for an experience requirement in an IT Job in a US City. An average salary for a new hire is calculated.

The standard deviation and t-statistics are applied to derive the salary distribution above and below the average. The highest, above average and below average salaries are specified for the information technology job market.

Market Salaries

The salary structure of the entire job market for an IT Job in a US City is presented in the Salary Report.

= and