A very wide variety of IT skills is needed to support an enterprise as diverse as state government. Each employee usually needs to have command of a suite of skills to accomplish daily work assignments, but no single employee would be expected to be proficient in all 126 skills assessed in the survey. In this study, employees assessed their own skill proficiency by rating themselves using a five-point scale that ranged from “none” to “expert.” Respondents were asked to reflect on their ability to apply each of the 126 skills in a work environment based on their personal experience and training, even if they did not currently use a particular skill. The proficiency scale and definitions are listed below:

• None – not able to apply this skill
• Basic – able to handle only the simplest assignments or tasks
• Intermediate – able to handle independently many types of assignments or tasks
• Advanced – able to handle independently nearly all types of assignments or tasks
• Expert – able to handle independently all types of assignments or tasks and serves as a role model or coach for others

Given the number of important variables that influence the state IT enterprise and workforce, no absolute measure of statewide proficiency is possible or appropriate. The skill proficiencies vary widely according to several factors. For example, some skills represent fundamental knowledge areas, such as principles of programming, which are prevalent in all agencies and most jobs. Others are very specific IT tools that are used only in certain situations or organizations, such as Java Studio. In addition, proficiency for each skill varies according to different contextual factors such as job specialty or because a certain agency uses one tool while other agencies use another for similar work. Moreover, the desirable distribution of proficiency levels can vary widely depending on these same factors. For example, for a very high-end technical tool, it may be quite appropriate that a few people have expert proficiency while most employees have little or none. For a different skill, one that is more linked to the essence of IT work such as principles of operating systems or system design and development, we would expect to see a wider distribution of proficiency with some employees at the high and low ends and most people in middle.

Because the proficiency ratings were based on a scale, we were able to calculate mean proficiency ratings for each skill. For purposes of estimating mean proficiency ratings, we omitted responses of “none” for all analyses except those which look at the data by job specialty. No single employee would possess proficiency in every skill. For example, we would expect very few people to report a proficiency in Linux because few state agencies have adopted or explored it. Similarly, a person whose job concentrates on desktop computing may have no need for proficiency in network architectures or mainframe operations. Therefore throughout the report (except for presentations that are based on job specialties) we use a proficiency rating scale of 1 – 4 where 1 = basic and 4 = expert. When exploring job specialties, however, we use the full five-point rating scale where 1 = none and 5 = expert. In these instances we included responses of “none” in our analysis to more accurately capture those instances where employees lack proficiency in skills that are actually relevant to the tasks and work activities they need to perform.

Table 5 illustrates the general proficiency rating data available for each skill. It shows the ten technical skills with the highest mean proficiency rating among employees who report having at least basic proficiency. (Table E4 in Appendix E shows similar data for all skills in the survey listed in alphabetical order.)

Table 5. Top ten technical skills by mean proficiency ratings

Skill	N	Mean*	Basic	Intermediate	Advanced	Expert	None
Principles of programming	2,588	2.36	566 (22%)	645 (25%)	722 (28%)	329 (13%)	326 (13%)
Windows operating systems	2,809	2.35	544 (19%)	926(33%)	916 (33%)	283 (10%)	140 (5%)
COBOL	2,781	2.21	589 (21%)	424 (15%)	453 (16%)	249 (9%)	1,066 (38%)
Principles of operating systems	2,759	2.18	710 (26%)	922 (33%)	729 (26%)	217 (8%)	181 (7%)
Support for desktop applications	2,793	2.13	669 (24%)	666 (24%)	506 (18%)	211 (8%)	741 (27%)
Hardware & maintenance support	2,795	2.09	710 (25%)	559 (20%)	477 (17%)	207 (7%)	842 (30%)
Help desk activities	2,803	2.05	792 (28%)	627 (22%)	497 (18%)	196 (7%)	691 (25%)
Systems implementation	2,791	2.04	717 (26%)	822 (30%)	498 (18%)	157 (6%)	597 (21%)
System life cycle planning principles	2,793	2.02	759 (27%)	707 (25%)	506 (18%)	145 (5%)	676 (24%)
Testing & evaluation	2,797	1.98	791 (28%)	866 (31%)	454 (16%)	151 (5%)	535 (19%)
Structured system analysis and design principles	2,793	1.98	743 (27%)	749 (27%)	457 (16%)	125 (5%)	719 (26%)

* Means include those employees who have at least basic proficiency, calculated on a four-point rating scale: 1 = Basic, 2 = Intermediate, 3= Advanced, 4 = Expert. Row totals may not equal 100% due to rounding.

Most of the top skills are broadly distributed across the levels of proficiency. For example, few employees reported having no proficiency in principles of operating systems, while the numbers reported at basic, intermediate, advanced and expert levels resemble a normal distribution with small numbers at the high and low ends, and most in the middle range. COBOL has a different rating distribution because its use is phasing out in most places and many employees have never used it. However, the mean proficiency rating among those who do have the skill is relatively high.