Thursday, January 29, 2015

Women in Engineering

By: Gary J. Salton, Ph.D.
Chief: Research and Development
Professional Communications, Inc.

An analysis of National Science Foundation data shows that the low participation of women in engineering is the result of systematic bias that exists across all areas of engineering. The cause of this phenomenon is investigated using a sample of 2,855 practicing non-management professional engineers. These were contrasted with 619 professional non-management scientists.

The analysis of the sample data traces the lack of female participation to information processing strategies used by the male participants in the two fields (i.e., engineering and science). These strategies are in turn dictated by the structural requirements of the two areas.

The scientific structure allows female’s to effectively use their preferred information processing strategies. This makes science a more attractive career destination for women.  The methods used to address the structural challenges of engineering tend to limit the female’s potential contribution. This makes engineering a less attractive area. The result is a participation rate half of that realized in the exact sciences.

The engineering impediment is not inherent. It is a social construction. Alternative methods that maintain the integrity of engineering while admitting the contributions of the entire range of information processing approaches—including those favored by current female participants—are possible.

Click for Video
A companion video leverages the ability of animation and visual descriptors to augment the explanations and analysis offered here. You can access this YouTube video from our website at or by clicking the icon on the right to go directly to the YouTube video.

Graphic 1 shows that engineering has consistently ranked at the lowest level of female participation in the STEM occupations. 

Graphic 1

The exacting demands of engineering are often cited as the major reason for the low female involvement. This is tested by comparing the female participation rate in engineering to that of the exact sciences. Table1 shows the result of that comparison (see footnote #1 for citation)

Table 1

The red figures in Table 1 show that the exact sciences have twice the level of female participation as does engineering. The fields are equivalent in complexity and importance. The level of rigor demanded is comparable. The comparison demonstrates that the shortfall cannot due to intellectual demands or female biochemistry.  Something else is at work.

Table 2 sorts the National Science Foundation list of 154 STEM occupations by level of female participation. The thumbnail labeled “engineers” shows female engineers clustered in the lowest levels of participation in STEM occupations. The thumbnail labeled “scientists” shows female exact scientists broadly distributed across the range.  The clustering confirms that whatever is causing the shortfall is systematic and applies to all areas of engineering (see footnote #2 for citation).
Table 2
National Science Foundation List, 2010
Table 3 lists some of those most cited “reasons” for low participation. To these can be added personal qualities such as assertions that engineers have low self-esteem, insecurity,  poor communication skills and other such negative personal attributes (See footnote #3 for source citation).

Table 3

All of the above “reasons” are true in individual instances. But none appear sufficient to account for the systematic bias evident in the Table 2 clustering. And the shortfall of females has consequences.  The things that could have been created or improved with female participation have no gender (e.g., Grace Hopper-inventor of COBOL; Stephanie Kwolek-Kevlar; Gail Boydston-Insulin production; Barbara McClintock-Nobel Laurite in Cytogentics, Marie Curie-dual Nobel Laurite in radioactivity; etc.--see footnote #4 for source and added citations). Everyone—male and female—loses . 

There is a financial as well as societal loss.  Females are over half of the population and can account for even more of a firm’s customer base.  A woman’s perspective in design, development and production would lead to refinements that improve the appeal of product offerings. The relative absence of women has a bottom-line cost.

Whatever that cause of the shortfall we know with certainty that the basic information processing model (input-process-output) must be involved.  “I Opt” Technology measures information processing.  Unique combinations of input and output elections produce characteristic behaviors called “styles”.  Table 4 provides a thumbnail outline of these styles (see footnote #5 for reference).

Table 4

The “I Opt” paradigm can be trusted.  It has been validated along all eight validity dimensions. It has also been subject to extensive “Stress Tests” (see footnote #6 for validity detail). The tool offers a reliable lens with which to quantitatively assess the engineering culture.

Any ratio scale can be restated as ordinal (i.e. rank order) categories. These measure the frequency with which a behavior will be seen. Favoring a particular approach even by a small amount will cause it to be the first deployed in addressing an unfamiliar issue. Observers note this frequency and incorporate it into the mental image of a person or group

Styles are short-term responses to particular situations. However, no style is applicable to everything. When a favored style does not apply people shift to the next most favored rank ordered style.  The rank order of styles is stable. Thus sequences of style use will tend to repeat. These repeated sequences are seen as long-term behavioral tendencies. These regularities are called Strategic Patterns and are listed in Table 5. They can temper or magnify shorter-term style based judgments (see footnote #5 for added detail).

Table 5
Long-Term Behavioral Sequences


The ability to measure the basic information processing model gives us a tool.  The availability of three distinct measures allows the use of the tool in assessing the engineering environment

The base sample totaled 2,855 working professional engineers who have no supervisory responsibilities. Of these 2,335 were male and 520 female (18.2%). The sampled engineers were drawn from 210 unique organizations. These included non-profits, government bodies and for-profit firms. Subsidiaries were consolidated into the parent since they are likely to be governed by the same policies, procedures and programs.  A majority of the engineers are based in the United States but there is significant international representation from 19 other countries (see footnote #7 for country listing) . 

A sample of 619 working, non-supervisory scientists was used as a tool of comparison and contrast. This included 423 males and 187 females (30.2%) drawn from 76 unique organizations (see footnote #8 for additional detail).

A reasonable question is whether females who choose engineering process information differently than do females who choose science. A divergence could be a basis for the difference in participation rates.

Graphic 2

Graphic 2 shows that women engineers are 9.5% more inclined to use the disciplined action of the Logical Processor style than are female scientists.  This difference is statistically significant at the marginal 5% level (p<.05). This condition is magnified when the focus changes to style frequency as is shown in Graphic 3.

Graphic 3

Frequency is determined by the rank order of the styles.  In this case over 20% more of the female engineers use the more stringent, action oriented LP style as their primary (i.e., first choice) way of addressing undefined issues (p<.05--see footnote #9 for significance detail).
Short-term responses are not the whole story.  People work together for extended periods. They witness behavior over multiple situations.  “I Opt” measures this long-term response sequences with its Strategic Pattern variable (see footnote #10 for elaboration on pattern measurement).

Graphic 4

A glance at Graphic 4 is enough to suggest the gender commonality in science and engineering. Calculations confirm that is no difference at any level of statistical significance.

The foregoing analysis shows that female engineers and scientists are drawn from the same talent pool.  There is a discernible difference in their initial responses (i.e., style). But that position is not held with any undue tenacity (i.e., style strength). Long run patterns cancel out these modest style effects. The differences between female engineers and scientists are real but without consequence.

Women in engineering and science are roughly similar in strategic posture.  A constant cannot be used to explain a difference. This suggests that the cause of low engineering participation might be found among the males.

Graphic 5

Graphic 6

Graphic 5 shows that male engineers are 10.6% more committed to the analytical HA style while male scientists favor the idea generating RI style by 13.1%. The frequency of display shown in Graphic 6 magnifies this difference. Both strength and frequency differences are highly significant (p<;.001).  

Graphic 7

The differences persist in the male’s long-run patterns. Graphic 7 shows that male engineers put 11.2% more reliance (p<.001) on the Conservator pattern (detailed analysis coupled with precise execution and a reliance on proven methods) than do males scientists. This pattern puts high value on certainty and sees failure as a condition to be avoided. 

Scientists register a 55.2% difference (p<.001) in the Changer pattern (new ideas quickly but tentatively implemented). The Changer pattern actively uses error as a guidance mechanism. Failure is accepted as a way of discarding less than fruitful approaches. For the scientist failure is a tool.

This brief analysis is enough to demonstrate that male engineers and scientists use different strategies.  They both draw on the same tools (i.e., math, scientific method, explicit standards, etc.).  But they employ them differently.  This difference is sufficient to create a cultural environment which can effect female participation.

The engineer’s favored strategies (HA style and Perfector pattern) rely on logical cogency, accuracy and comprehensive understanding.  These factors can be tested with rational inquiry. Rational challenges are a way of minimizing the risk of failure. The need to use convincing rationality makes divergent perspectives more expensive. For example, there is little common ground between options based on expediency and those based on optimality. The need to consider efficiency and effectiveness compounds the problem. The net result is that small divergences can carry high cost. Negative stereotypes become a low cost way of dismissing variant perspectives.

Scientists also rely on the analytical HA style as a short-run posture but with less intensity. They are likely to be more willing to entertain other options.  And their strong secondary idea-oriented RI accepts incomplete specification as the natural condition of new ideas. Their experiment-oriented Changer pattern relaxes the need for failure avoidance.  The net result of the scientist’s orientation is an ability to use a greater variety of information postures. 

Science and engineering use the same tools in different social structures.  The ability to utilize female skill sets depend on how receptive these structures are to the female’s information processing orientation. To find that out we need to compare the males and females in each discipline.

Graphic 8 compares the style strength of male and female populations in both engineering and science. 

Graphic 8

The magnitude of difference in the two groups can be gauged by summing the absolute difference (i.e., disregarding the direction of difference) of each style in the profile. The engineers register a cumulative 24.0% strength difference while the scientists register 27.8%. The difference is small. The important distinction is in where that difference is happening.
Female scientists are more committed to the analytical HA (8.6%: p<.05) while males favor the idea generating RI styles (14.1%: P<.01).  These are complementary styles.  Both appreciate the value of new alternatives. Interactions are likely to be seen as mutually supportive—males proposing more ideas and females refining them through analysis.
Engineering is a different story. The differences are in the same order of magnitude (10.2% in LP and 11.6% in RI styles both at p<.001 level). But the LP and RI are contradictory positions. The LP’s preference for proven, well-understood methods forecloses the RI’s favored preference for new, untested ideas. Male-female engineer interactions are thus less likely to be mutually supportive.

 Graphic 9 

Graphic 9 shows that the frequency of style elections have the same complementary (science) and contradictory (engineering) relationship as was seen in style strengths. Any judgments based on styles are reinforced. 

 Graphic 10

Graphic 10 shows that females in science tend to be rely more heavily on the Conservator pattern (proven methods systematically applied) while male scientists tend to put more stress on the Changer pattern (new ideas experimentally applied). This introduces a note of gender based tension in science (both differences are statistically significant at  p<.05). However the two remaining pattern dimensions are aligned thus confining the scope of any tensions. 
Engineering gender distinctions are much more pronounced and wide-spread.  Graphic 10 shows that female engineers differ from their male counterparts on all four pattern dimensions (all are statistically significant at the p<.05 or better). This broadly reinforces any negativity generated by the short-term style postures. Engineering offers much more fertile ground for negative stereotypes to take hold.

Females in both science and engineering are differ significantly from males. However, females in the sciences differ in a more complementary manner. Females in engineering are likely to have a more contentious relationship with their male counterparts. Stereotypes are more easily generated in more difficult environments.
The difficult character of female participation in engineering effectively restricts their ability to contribute. The loss of value is not recognized because the male approach works “well enough.” It is rather like evolution. The creature produced is not optimal but just sufficient relative to the environment within which they exist.

Females comprise over half of the population, technology has eliminated many physical and biological constraints (e.g., strength, childbearing, medical, etc.) and women have demonstrated intellectual capacity equal to men.  Clearly, there is an advantage to be gained from using these assets in the engineering profession. Realizing this advantage will require that actions be taken.

The current condition of females in engineering is governed by social factors (see footnote #11 for elaboration). Social causes reside in the relationships among group participants. The typical character of those relationships is directed by the conditions of survival.  For engineering the dominant survival condition is risk minimization.

The current method of controlling risk is adversarial. Engineers test each others positions using intellectual tools –discussion, debate, disputation, etc.  For this strategy to work positions must conform to a common logical framework.  For engineers that framework is dictated by the structured logic of the Hypothetical Analyzer style. Deviations from that style (e.g., appeals to authority, expedient options, incomplete ideas, etc.) are summarily dismissed or depreciated.

This analysis demonstrated that females currently in engineering favor strategies significantly different from those of the males.  This means that divergent behaviors will tend to repeat. Generalizing this commonality into stereotypic expectations is a predictable outcome. There is no single “silver bullet” solution. It is likely that a mix of initiatives will be required.  These might include:

Encouragement: Maintaining the current efforts to encourage females in STEM education is essential. The more qualified females, the greater the numbers likely to elect engineering as a career. An increasing number of females will re balance the dominant majority. At some point viability of any stereotypes will become untenable. 

Profile Mix: Efforts to attract women with greater RI (idea generating) capacities also merit pursuit.  Success would better align the female profile with those of the male. In addition, the RI style is a key to both formal and informal leadership roles (See footnote #12 for reference).  More female leaders could help to provide female role models while also serving to dismantle negative stereotypes.

Attracting “creative” females will take more than marketing appeals.  People with higher RI capacities are attracted by the opportunity to exercise this capacity. They are looking for the chance to make major contributions.  Highlighting the opportunities available in the various branches of engineering could help capture the interest of this group.

Compensation Structure: Females are intellectually and biochemically as equipped as males to fulfill the demands of professional engineering.  But there are differences.  Societal roles can cause women to value various components of the compensation structure differently than do men. Engineering can make itself more appealing by designing compensation to better accommodate females.

One method of doing this is to allow the employee—male and female—to allocate total rewards between base salary, incentives and fringe up to a market based total compensation level (See footnote #13 for illustrative listing). This strategy disadvantages no one while positively leveraging any role based value differences.

Physical Environment: The popular open office concept creates an opportunity for inappropriate attributions, casual offensive banter and other gender hostile behaviors. Highly interactive environments can easily breed hostility and dysfunction.  Providing offices or partitioned areas would automatically minimize the occasion for hostile occurrences. In addition, the improved environment is likely to be welcomed by all involved—male and female.

Knowledge: Intellectual tools are as much a part of structure as is the physical environment. An earlier study showed that the Engineering Personality (See footnote #14) arose from predictable behaviors governed by the structure of the work. The predictability generated by the information processing profiles of current female engineers is at least partially responsible for negative female stereotypes. But there is a difference.  

The predictability of the Engineering Personality is governed by the laws of nature inherent in the materials and methods being used. The predictability of social interactions is governed by the interpretation of the meaning and value of the social content. Change the meaning and the social outcome changes. 

“I Opt” is a tool designed from the onset to affect social situations. It has used for over 20 years by some of the largest firms on a worldwide basis to positively affect teams and guide leadership behavior  (See footnote #15).  Since it is based on engineering concepts and is grounded in quantitative measurement it is readily accepted by professional engineers.  The basic “learning” provided by “I Opt” involves how to use the existing profiles of group members to best further the common interest.  No one has to change and everyone ends up better off.  “I Opt” is a tool that merits consideration in both corporate and university settings.

The foregoing ideas are illustrative rather than exhaustive. They involve changing the structural environment and/or equipping people with skills necessary to favorably navigate their social environment. However, in final analysis the issues involving gender based impediments will be resolved when female participation reaches a tipping point. Even then issues will remain.  But they will be local rather than systemic. Given the potential gains for all involved, the game is well-worth the candle.

(1) Scientific areas which had complexity and precision demands similar to engineering were extracted from the National Science Foundation, National Center for Science and Engineering Statistics, Scientists and Engineers Statistical Data System (SESTAT) (2010),

The table used was Appendix Table 3-13, Employed scientists and engineers, by sex and occupation: 2010.  It was retrieved September 27, 2014 from

The  National Science Foundation lists areas such as Sociology, History and Accounting among the sciences.  For purposes of this research only those professional areas which demand rigor comparable to engineering were extracted as representing the exact sciences.


(2) The thumbnails were imaged from a reformatted Microsoft Excel listing published by the National Science Foundation, National Center for Science and Engineering Statistics, Scientists and Engineers Statistical Data System (SESTAT) (2010), . The file is most easily retrieved at The textual narrative can then be searched for the link labeled (appendix table 3-13)” to obtain the excel worksheet. 

(3) Moore, Susanne (2012), Dealing with the ‘Boys Club’ .A blog entry self-cited as “one of the most read whitepapers on my previous company website and still resonates today.” Retrieved October 10, 2014: 

(4) Washington State University (2004), Famous Women Engineers, Retrieved September 29, 2014  

(5) Salton, Gary (2008). "I Opt" Strategic Styles and Patterns. watch?v=KVOyznCCWB8. This video provides an orientational overview. A complete explanation of the theory and application of the technology is available in a ~17 hour online/telephone counseling certification program conducted on demand by Shannon Nelson ( 

Strategic Patterns are predictable sequences of style elections.  Table 5 shows the common combinations.  Two additional combinations labeled “Split Styles” are omitted for purposes of clarity. These are relatively infrequently encountered in practice and are covered extensively in the Certification Program. 

As noted in the text, strategic patterns are measurable in both strength and frequency. Pattern frequency is relatively easy for an observer to discern over a series of transactions. Pattern strength is more difficult to recognize and thus has a lesser effect on an observer’s evaluations. For purpose of this study, the pattern strength effect has been omitted. 

(6) “I Opt” Validity: The basic validity study was published in book form by HRD press in 2001. The book can be purchased from Professional Communications or read free of charge in pdf format on its website at: The academic citation for the book is: 

Soltysik, Robert (2000), Validation ofOrganizational Engineering: Instrumentation and Methodology, Amherst: HRD Press. 

Additional elements of validity verification can be seen embedded in an extensive number of formal studies published in both textual formats at: and Many of the studies are also available in video format at:

The Strategic Style “Stress Test” focused on short-term decision-making can be found in textual format at or in video at 

The Strategic Pattern “Stress Test” focused on long-term decision-making can be found in textual format at or in video at 

(7) The countries included in the engineering sample are: 

(8) Scientists were drawn from 76 unique for-profit, government, military, universities and non-profit organizations. Job titles as assigned by the firms, reported by participants or cited in published sources were used for classifying a person as “scientist.” 

(9) Significance tests for style strength (a ratio denominated variable) were determined using the Microsoft Excel t-Test: Two-Sample Assuming Unequal Variances statistic. Frequency is measured using ordinal data. Rank ordered data are best tested using a non-parametric statistic. In the case of this study ordinal data were tested using Virginia Commonwealth University’s online Mann-Whitney U test at 

(10) Strategic Patterns can be measured in terms of strength (i.e., ratio) as well as frequency (i.e., rank order). For purposes of this study pattern strength is only marginally relevant. In the interests of clarity and brevity it is not addressed. 

(11) The greater participation rate of female scientists (vs. engineers) suggests that any bias is not inherent in the male gender. The commonality of females in both fields eliminates biology and chemistry as causes. The rough equivalence of the intellectual tools used in both fields allow us to any dismiss cerebral limits as a cause. The fact that the samples were drawn from the same population means that general culture cannot be the reason. Any psychologically based cause would require the proponent to identify some specific quality that differentially attracts “pig headed, insensitive, workaholic, etc.” men to engineering and not science. Thus biology, chemistry, culture and psychology can be dismissed as probable causes. 

(12) Salton, Gary (2012). Organizational Rank and Strategic Styles. This study used a sample of 10,617 individuals from 1,559 different organizations to show a systematic relationship between the RI strategic style and hierarchical rank. It also isolated a single reason why this relationship exists and why it will continue to persist into the future. The research blog can be read at 

A companion YouTube video that both abbreviates and expands on this research blog can be viewed at: 

Additional Evidence-Based Research Blogs showing systematic relationships between strategic style and organizational rank include: 

Salton, Gary (November 2010) Sales Management and Performance. 

Salton, Gary (October 2010) City Management 

Salton, Gary (September 2009). The Nursing Staircase and Managerial Gap 

Salton, Gary (September 2008). Hierarchy Influence on Team Leadership 

Salton, Gary (August 2008). Engineering Leadership. 

Salton, Gary (June 2008). The Pastor as a Leader. 

Salton, Gary (May 2008). Fitting the Leader to the Matrix 

Salton, Gary (October 2007). Leadership, Diversity and the Goldilocks Zone 

Salton, Gary (October 2007). How Styles Affect Promotion Potential 

Salton, Gary (November 2006). Gender in the Executive Suite 

Salton, Gary (October 2006). CEO Insights

(13) Below is an illustrative listing of some of the items which could be priced and incorporated into a person’s total compensation. Central to this idea is that tradeoffs involve all forms of compensation including the base wage. 

BASE SALARY                                                $$$$$$$$$ 
Childcare                                                         $$$
Community Involvement Support                         $
Defined Benefit Retirement Plan                         $$$$
Defined Contribution Retirement Plan                  $$
Dental Care                                                      $$
Dependent Care Reimbursement Account           $$
Employee Assistance Programs                         $
Financial Planning                                             $
Flexible Spending Accounts                               $$
Flexible Start/Quit Times                                   $$
Flexible Workplace                                           $$
Health / Medical Care                                        $$$$
Health Savings Account                                     $
Healthcare Reimbursement Account                   $$
Holiday Pay                                                      $
Life Insurance                                                   $
Long Term Care Insurance                                  $
Long Term Disability                                           $$
Non-Production Bonus Pay                                 $$
Onsite Daycare                                                  $$
Paid Family Leave                                              $
Paid Funeral Leave                                             $
Paid Jury Duty                                                   $
Paid Military Leave                                              $
Paid Personal Leave                                            $
Paid Sick Leave                                                  $
Payroll Deduction IRA                                         $
Prescription Drugs                                               $
Retiree Health Benefits                                        $
Retirement Annuities                                           $$
Short Term Disability                                           $
Stock Options (Performance)                                $
Subsidized Commuting                                        $
Tele-Work / Telecommuting                                  $
Unpaid Family Leave                                          $$
Vacation Pay                                                     $$
Wellness Programs                                            $ 
TOTAL COMPENSATION                                    $$$$$$$$$ 

(14) Salton, Gary (2014). The Engineering Personality. This research can be viewed in textual format at 

It is also available in video format on YouTube at:

(15) “I Opt” technology has been used by universities and corporations in both classroom and workplace settings. In classroom settings the technology is typically used educational tool to provide a general framework for understanding human interaction. A variety of reports addressing issues such as self-discovery, learning, career and sales are used to support these activities. 

In workplace settings “I Opt technology is used to improve performance of existing groups and teams. The tools typically used in this context include: 

The “I Opt” TeamAnalysis™ report shows a mechanism for analyzing a team using quantitatively based methods especially appealing to the exacting standards of engineers. It can be viewed as a video at: 

The “I Opt” LeaderAnalysis™ report views a group from the perspective of the leader in a way that considers the opportunities and exposures inherent in the specific group of people being led. This video offers extensive explanations of analytical technique as well as various supporting graphics and tables that are appealing to engineers. The video can be viewed at:

The “I Opt” Emotional Impact Management™ report focuses on controlling the emotions that a person causes through their interactions. Positive and negative emotions directly affect the other person’s likely performance on matters of common interest. This video can be viewed at: