Observations on the National Academy Report on Sexual Abuse in STEM Departments

by Pat Frank

A few days ago, Charles TM posted an essay about the recent National Academy Report, “Sexual Harassment of Women: Climate, Culture, and Consequences in Academic Sciences, Engineering, and Medicine,” [1] which can be found at WUWT here.

Apparently, the NAS Report is so outspoken about widespread abuse of women in academic science, technology, engineering, and mathematics (STEM) departments that Senators Kamala Harris, Jacky Rosen, and Richard Blumenthal plan to introduce legislation to effect a rescue.

The NAS Report purports the case that not only is the sexual abuse of women wide-spread, but also that the patriarchic hierarchical structure in STEM departments cultivates an intersectionally abusive environment where sexual harassment of women becomes likely.

Before proceeding, we need to acknowledge that there have been instances of sexual abuse in the STEM academy that were widely discussed. Well-publicized examples include one at Berkeley in 2015, one at Yale in 2014, one at CalTech in 2016, one at the University of Chicago in 2016, one involving three women at the University of Alabama in 2013, one at Arizona State University in 2015. There are serious episodes involving real abuses and real victims.

The issue, however, is not about episodic abuse, which is pretty much inevitable anywhere there are large numbers of men and women working under one roof (I get to this issue below. Abuses arise with the appearance of low-probability personality types in statistically valid populations).

Rather, the issue is about whether the data support a conclusion that sexual harassment of women is wide-spread in academic STEM departments and that the environment in STEM departments cultivates the sexual harassment of women.

The NAS Report came to roost at Stanford in October last year (2018 if you’re a reader from the distant future). After giving it a look-through, I decided to investigate the same question as I did in 2001 when the IPCC TAR came out: are the accusatory polemics justified?

Well, guess what. …

This will be a summary. The full assessment is 30 journal-like pages plus five Figures, two Tables and 139 citations; long and involved.

So, here’s the summary:

· The NAS Report traverses from wrong to meaningless.

· It misrepresents the literature.

· It misrepresents its own data, purveying false and inflated rates of sexual harassment.

· It is a product of almost unprecedented scholarly incompetence.

· It is calculated to stampede universities into a false moral panic.

· It is the opening barrage of the academic Cultural Studies warriors’ war to wreck science.

Those who wish the Full Monty can download a pdf for themselves here (876 kB). Please choose the “slow download” option. It’s fast enough and you’ll avoid pecuniary offers. Also, the file has no viruses.

If you read and like it, please feel free to pass it along to wherever or whomever you think it might do some good.

The full assessment critiques the survey methodology of the NAS Report. It examines the personality aspects of sexual harassment. It introduces and literature-validates the sexual harassment males by females, by their peculiar power of sexualized display in inappropriate venues.

The full assessment also discusses a widespread and generalized defamation of males by a foolish statistical inference that falsely paints every male as a potential abuser.

The last part diagnoses the NAS report itself as the opening attack of Cultural Studies warriors war of aggression on science itself. They intend its destruction. The final paragraphs here return to this.

This post just summarizes the problems with the NAS Report itself, and then moves into who actually sexually abuses.

Part I: The National Academy Report on sexual harassment in STEM fields misrepresents or omits critically central scholarship, and mangles its own analysis.

The Report relies upon sexual abuse statistics generated by a large-scale survey on the campuses of the University of Texas system. The survey used a version of the “Sexual Experiences Questionnaire,” (SEQ).

The SEQ was developed by Louise Fitzgerald in 1988. [2] The UT survey was a modification of a second version of the SEQ published in 1995. [3]

The SEQ version used by the NAS Report was yet a third version and had four sections. Each section asked a set of four questions about one level of sexualized experience. Each succeeding section (factors, in the parlance), asked about increasingly personal interactions.

Here’s the set of four questions for the first factor of the NAS SEQ, for example:

1. Sexist Hostility/Sexist Gender Harassment: Unwanted and unwelcomed words, actions, symbols, gestures, and behaviors that are based on sex or gender and characteristically repetitive.

1.1. Treated you “differently” because of your sex.

1.2. Displayed, used, or distributed sexist or suggestive materials.

1.3. Made offensive sexist remarks.

1.4. Put you down or was condescending to you because of your sex.

The second set of factors asked about, “Sexual Hostility/Crude Gender Harassment.” The third set was, “Unwanted Sexual Attention,” and factor set four was, “Sexual Coercion.” None of the factors asked about violent sexual abuse or rape.

The 1988 and 1995 versions of the SEQ ended with, “Have you ever been sexually harassed?” (1988 version) or the statement “I have been sexually harassed” (1995version). These are called the criterion question or statement.

Positive answers to the criterion question permitted concluding that the SEQ behaviors added up to harassment. Negative or blank criterion answers implied the respondent didn’t feel harassed by the experiences.

The meaning of each SEQ question is open to the sensitivity of the respondent.

For example, 1.1 could be answered ‘yes’ if a male held a door open for a given woman. Some women take umbrage at that. Others appreciate it. 1.2 might be a sex-joke birthday card. 1.3 might be taking offense at, ‘men are such beasts!’ 1.4 might be, ‘that stuff is too heavy for you to carry alone (said by someone of one sex to someone of the other).

These examples are all innocuous of course, but nevertheless each one positively meets the requirement of the question. The questions are all very ambiguous and have no context. Individual sensitivity to umbrage-taking varies.

As I’ll show later, there are sex-related differences in personality. Merely mentioning sex-differences to someone could energize offense-taking under any and all of the factor one questions.

This ambiguity was among the very serious criticisms Prof. Barbara Gutek and her colleagues made of the SEQ back in 2004. [4]

Here’s a summary of their critical findings concerning the SEQ:

1. The SEQ is not strong in accuracy or precision.

2. Test, re-test data are poor.

3. The causal direction of harassment is left ambiguous.

4. The SEQ is not standardized. There are several variants. SEQ results cannot be trended across time or space.

5. The SEQ over-reports the prevalence of sexual harassment.

6. The SEQ wording does not allow one to conclude any respondent was personally overwhelmed or felt threatened by the experiences.

7. A far larger fraction report SEQ experiences than answer the criterion question/statement that they felt sexually harassed (a big problem for researchers looking to butter their academic bread with sexual harassment.).

8. Cross correlation between the SEQ factors is about 0.72.

The causality criticism of item 3 is fundamental. Does a given SEQ score mean that a harassment-tolerating environment caused the sexualized work behaviors? Or does it mean sexualized work behaviors caused the respondent to infer a harassment-tolerating environment?

The former is the researcher favorite, because it allows them to conclude that social construction causes sexual harassment, rather than that harassment comes from individual behavioral choices. The social construction bad-boy of choice is the patriarchal hierarchy; very convenient if one makes a career as a cultural studies warrior.

Gutek, et al., concluded that, “the SEQ is a flawed instrument and that its positive features have been greatly exaggerated. It does not seem to measure anyone’ s definition of sexual harassment, including that of its own developers.”

They also quoted a judge who rejected SEQ results in concluding a sexual harassment case in 2002:

“In EEOC v. Dial Corporation, Nov. 17, 2002. Ill, No. 99 C 3356, the Federal judge observed that the, “survey instrument presents inherent reliability problems” and “the SEQ portion of the survey lacks validity” … “This lack of comparability of SEQ scores seems problematic to me, because it seems to render the SEQ scores devoid of any objective meaning” (p. 8). The judge concluded that the “survey materials” [i.e., the SEQ — P] are too flawed to be useful in assisting the fact finder in this case.” (my bold).”

The judge came to the same conclusion that Gutek et al., did: the SEQ does not measure sexual harassment. In fact, it seems to have no objective meaning at all.

And yet, this utterly flawed SEQ survey is at the center of the NAS Report; chosen by professionals in the field to objectively reveal the hidden abuse within academic STEM departments.

Correlation raises its head: Item number 8 in the Gutek, et al., list of flaws is particularly acute.

The four separate SEQ factors were originally said to measure four separate dimensions of behavior. That means each of the four factors is supposed to be orthogonal to the others. That means they are 90° apart in SEQ phase-space; ideally with an inter-correlation coefficient = 0.

However, the factors are not uncorrelated. They are correlated, and their correlation coefficient averages 0.72. Correlated factors do not have a unique meaning.

Correlated epidemiological data sets must be corrected by a factor of 1/(1+r), to obtain the number of statistically independent data points (“r” is the correlation coefficient). [5]

However, the NAS authors overlooked this step in their analysis. The correction to the SEQ data is 1/(1+0.72) = 0.58. All the fractions of respondents who had SEQ experiences must be multiplied by 0.58 to find the statistically unique fraction.

So, for example, the Report makes a lot of hay about the fraction of women who report SEQ experiences in the STEM fields.

They report that 22% of females in non-STEM academic fields, 20% of females in Science, 27% in Engineering, and 47% in Medicine, have all had at least one SEQ experience and therefore have been sexually harassed.

Those fractions must be corrected by 0.58 scaling, which yields, 13%, 12%, 16%, and 27%, respectively.

The Figure 2 bar-graph shows the raw SEQ scores and their statistical corrections.

Another sign of professionalism among the NAS authors is that they left the criterion question off their survey. This allowed them to promote the SEQ experiences themselves into instances of sexual harassment. This is a knowing misrepresentation of the SEQ results.

Here’s the NAS definition of sexual harassment (p. 28):

“Sexual harassment (a form of discrimination) is composed of three categories of behavior:

1. gender harassment (verbal and nonverbal behaviors that convey hostility, objectification, exclusion, or second-class status about members of one gender),

2. unwanted sexual attention (verbal or physical unwelcome sexual advances, which can include assault), and

3. sexual coercion (when favorable professional or educational treatment is conditioned on sexual activity).”

The three NAS criteria are actually individual SEQ behaviors that might or might not be harassment. Only the respondent could know. Context was important. Only the person experiencing the behaviors knew the context, and could judge whether they felt harassed, or not.

Originally, the SEQ experiences rose to harassment only if the respondent answered the criterion question as having felt harassed.

But the NAS authors have dispensed with the respondent’s perception. They now decide. They judge harassment from minimal evidence — a SEQ check-list — and without knowing any context. Any recorded SEQ behavior now counts as sexual harassment.

So, for example, they say that 20% of women in science were sexually harassed, not that 20% had SEQ experiences that may or may not have been offensive.

The NAS authors have falsified the meaning of the SEQ experiences. The result is to hugely inflate the fraction of women who can be said to have been sexually harassed.

The absence of meaning: Here’s another instance of professional incompetence in unaddressed factor correlation. The correlation coefficient is the cosine of the angle between result vectors in the data phase-space. [7]

For the SEQ factors, with correlation r = 0.72, that phase-space angle is cos^-1(0.72) = 44°.

Figure 2 illustrates the meaning of that phase-space angle. The meaning is that none of the SEQ factors have any unique meaning.

Result vectors that each have one unique meaning are uncorrelated in their data phase space. Their vectors are 90° apart. [8] When data vectors are separated by an acute angle, such as 44°, each vector has a projection on the opposed axes.

In a multi-dimensional phase-space, there are multiple orthogonal axes. Acute vectors will have projections on at least two axes, and possibly more than two of them.

In our physically real space, that means each SEQ factor is a mixture of two or more meanings.

In the SEQ, the meanings of the axes are not known. Even the dimensionality of the SEQ data phase-space is unknown, though it all has something to do with human behavior and socio-sexuality.

But the acute angle between the factor vectors means that none of the factors has a unique meaning. Each factor may have many meanings. And each one of the possibly several meanings is cryptic.

So, combining Figure 2 with the judgment of that perspicacious judge and the Gutek et al. critical analysis, it becomes clear that the SEQ results have no unique meaning at all.

The most condemnatory aspect of this whole thing is that the NAS authors knew of Gutek’s critical paper. I know that because Gutek, et al., is cited in some of their published papers (though without acknowledging the critical content).

But in the NAS Report they are utterly silent about, and neglectful of, Gutek, et al., 2004. They did not cite it in the NAS Report. It doesn’t exist in the NAS universe.

Blinded by the light: The NAS authors are very self-congratulatory, calling themselves a “committee of distinguished scientists, engineers, and physicians, and experts in sexual harassment research, legal studies, and psychology.”

And yet, they missed every single critical aspect of their work.

Of course, if they’d paid attention to those critical aspects, they’d have had nothing to talk about. The NAS Report authors have based their conclusions and recommendations on nothing definable.

All the faults, the errors, the oversights, and the negligence incline the same way: toward indicting STEM academics for sexual harassment. The NAS authors have calumniated tens and tens of thousands of innocent people.

The NAS Report is incompetence plus bias; all of it. It will do nothing but roil STEM departments with suspicion. It will poison professional relationships.

Attention government and lawyers: It looks to me like the NAS Report sponsors conditions that meet the EEOC definition of sexual harassment. The one that involves creating a hostile workplace by way systematic gender-directed harassment; harassment directed specifically against innocent males.

In my unlawyerly opinion, there’s a case to be made for a class-action law suit against the NAS for incitement and against any organization that imposes their views as authoritative. I believe evidence of real injury could be found in jobs lost and careers stunted. If so, the NAS authors might also be vulnerable.

Part II: Who abuses? Personality inventory has come a long way over the 100 years or so of its development. The HEXACO inventory divides personality into six general traits. These are Extraversion (E), Agreeableness (A), Conscientiousness (C), Neuroticism (N), Openness to Experience (O) and Honesty-Humility (H-H). [9, 10]

The average inter-factor correlation among these six is r = 0.11±0.08. That gives an average phase-space angle of trait-vector separation of (84±5)°; rather better than the SEQ.

Male and female personalities are distinguishable, [11, 12] much to the fury of social constructivists and gender-fluidophiliacs everywhere. The personality categories are not without a causal link to genetic inheritance and evolutionary biology. [13, 14]

Figure 3 illustrates some of the differences and similarities between the two sexes for two of the six HEXACO personality traits, plus the Barratt BIS-11 Impulsivity (I) trait.

People high in H-H and A are generally socially friendly. Too high gets into the friendly-to-a-fault category. So, in personality trait as in all things, more is not necessarily better. Impulsivity measures risk-taking or its lack.

Surveys show that people who are low in both H-H and A score high on the Likely to Sexually Harass (LSH) scale. [15, 16] The people, male or female, prone to sexually harass tend to have personalities low in modesty, straightforwardness, warmth, and kindness, and relatively high in rudeness and harshness.

The transition to low H-H or low A is found one standard deviation below the mean. I took the transition values for males to mark entry to difficult personalities for females as well. The female higher mean A indicates fewer females than males should have harsh personalities.

Assuming the H-H and A personality traits are heritably uncorrelated and independent, we can find out what fraction of people in a large population will have personalities low in both H-H and A qualities.

That fraction is just the product of the H-H and A Gaussian areas below the 2.74 value (Figure 2).

For males that’s 0.37×0.17 = 0.06, and for females it’s 0.41×0.04 = 0.02. These people all score high in the LSH scale.

So, the prediction is that 6% of males and 2% of females will have a proclivity for sexual harassment. The rest — 94% of males and 98% of females — will not.

This, by itself, is enough to indicate that sexual harassment is not a general character trait of males or females.

The claims of “toxic masculinity” and “rape culture” find no support in the psychometric literature. Both derogatory categories mounted against males are just self-serving tropes of Feminist “grievance studies.” [18]

Although I don’t discuss it here, the full assessment makes a very good case for harassment of males by females. That would be the 2% of females low in both H-H and A.

The way they harass males is by misuse of a female power very well documented in ancillary literature: by sexualized invitational displays of skin, but made in professional, pedagogical, or other serious venues. [19] It’s so common as to be considered normal.

Anyone interested in more on this topic is encouraged to download the full study and see for yourself whether you agree.

Finally: who offends violently? The literature pretty well establishes that people highly impulsive and low in H-H and A are of criminal tendency. So, high LSH and high Impulsivity combine to predict the fraction of the population that will offend through violent sexual abuse and rape.

The Impulsivity consistent with criminality is 1.58 standard deviations above the mean (the arrow in Figure 3, right). That turns out to represent 6% of the male or female population with high Impulsivity alone. But high Impulsivity alone does not predict criminality.

To predict criminality, we need low H-H and low A as well as high Impulsivity.

Again assuming trait independence, for males the Gaussian fraction is 0.37×0.17×0.06 = 0.004 and for females is 0.41×0.04×0.06 = 0.001. So, about 0.4% of males and 0.1% of females in a general population are predicted have personalities that will dispose them to violent crime, including to violent sexual assault.

That means 99.6% of males and 99.9% of females are not disposed to violent sexual crimes. Once again “toxic masculinity” is objectively disproved. Now, however, we can also objectively eliminate toxic femininity from the objectively determined list of characteristic human attributes.

Rape culture, already out the window, is now gone from the universe; except as a residual trope in what I’ve come to call ‘harpy culture.’

The prediction that 0.4% of males represents the fraction of violent offenders was independently confirmed by a 2015 study. I didn’t know about that paper before working through to the prediction, but happened across it while generally searching the violent crime literature.

In Quebec, Bouchard and Lussier were able to estimate the “hidden” population of violent sexual offenders, given a known number of convictions and re-convictions over a 42- month time window. [20]

They estimated 8,322 total violent male sexual offenders from 387 convictions and 9 re-convictions in a population of 2.13 million males aged 18-64. This yields a 0.4% violent sexual offender fraction among the male population. Confirmation of the prediction independently made from personality inventory could not have been more exact.

All these findings are well literature-validated in the full assessment.

Female STEMers are thriving: Direct surveys show that the overwhelming fraction of female STEM members (90-95%) are satisfied with their careers.

Females who want to be STEM workers face no obstacles in today’s US, and are even preferably hired. [21, 22] The disparity in percent of females in the math-intense jobs has been connected to their free choices for more human-centered careers. [22-26]

The more egalitarian a society, the greater the gender disparity in math-intense jobs. The reason for that is, when people are equal and free to follow their own interests, the biologically (evolutionarily) determined preferences are most able to play a dominant role in career choices.[27, 28]

There just is no validity at all to any of the charges made in the NAS Report; not one of them.

Females are doing fine in STEM careers. There is no crisis of academic sexual abuse.

The set of recommendations offered in the NAS Report is the fondest dream of the Academic Left: ‘Universities should impose equity hiring, and cultivate a politically submissive STEM nomenklatura overseen by social-justice commissars: a new Lysenkoism, in other words. [29]

I can’t imagine a better program to destroy science.

The people who wrote The NAS Report are professional academics specializing in sexual harassment. They not only mangled the literature, they applied it to a malign end.

They could not have exhibited greater incompetence.

A dire warning: Readers here may or may not be aware of the “Walkaway Movement.” Whether or not, you should know it generated a number of remarkable youtube videos.

A woman who moved away from Feminist ideology and found her way to free thought posted a 15:12 video.

She talks about her disenchantment with Feminism and its misandry, and her realization that hiring for equity is bigotry. She speaks from direct experience of the Humanities academy.

At 4:50 she gets to the intellectual and moral bankruptcy of Feminist and cultural studies theory, the aggressively anti-rational indoctrination agenda of the academic Humanities in general, and their consciously pejorative distortion of history.

At 15:12 she ends with, “Good luck sciences. They’re comin’ for ya. Pretty soon you’re not going to do studies on biological sex, because there is no such thing.”

I offer this: the NAS Report is the opening barrage of the academic Humanities in their conscious, deliberate, and malignant war to destroy science. They are an existential threat.

The Humanities have become the necrotizing fasciitis of academic integrity. The only cure for NF is early amputation.

References:

[1] National_Academies_of_Sciences_Engineering_and_Medicine, Sexual Harassment of Women: Climate, Culture, and Consequences in Academic Sciences, Engineering, and Medicine. 2018, Washington, DC: National Academies Press. 312.

[2] Fitzgerald, L.F., et al., The incidence and dimensions of sexual harassment in academia and the workplace. Journal of Vocational Behavior, 1988. 32(2): p. 152-175.

[3] Fitzgerald, L.F., M.J. Gelfand, and F. Drasgow, Measuring Sexual Harassment: Theoretical and Psychometric Advances. Basic and Applied Social Psychology, 1995. 17(4): p. 425-445.

[4] Gutek, B.A., R.O. Murphy, and B. Douma, A Review and Critique of the Sexual Experiences Questionnaire (SEQ). Law and Human Behavior, 2004. 28(4): p. 457-482.

[5] Hanley, J.A., et al., Statistical Analysis of Correlated Data Using Generalized Estimating Equations: An Orientation. American Journal of Epidemiology, 2003. 157(4): p. 364-375.

[6] Ilies, R., et al., Reported incidence rates of work-related sexual harassment in the united states: Using meta-analysis to explain reported rate disparities. Personnel Psychology, 2003. 56(3): p. 607-631.

[7] Rodgers, J.L. and W.A. Nicewander, Thirteen Ways to Look at the Correlation Coefficient. The American Statistician, 1988. 42(1): p. 59-66.

[8] Rodgers, J., A. Nicewander, and L. Toothaker, Linearly Independent, Orthogonal, and Uncorrelated Variables, Vol. 38. 1984. 133-134.

[9] Lee, K. and M.C. Ashton, Psychometric Properties of the HEXACO Personality Inventory. Multivariate Behavioral Research, 2004. 39(2): p. 329-358.

[10] McCrae, R.R. and O.P. John, An Introduction to the Five-Factor Model and Its Applications. Journal of Personality, 1992. 60(2): p. 175-215.

[11] Kajonius, P.J. and J. Johnson, Sex differences in 30 facets of the five factor model of personality in the large public (N = 320,128). Personality and Individual Differences, 2018. 129(p. 126-130.

[12] Verweij, K.J.H., et al., Individual Differences in Personality Masculinity-Femininity: Examining the Effects of Genes, Environment, and Prenatal Hormone Transfer. Twin Research and Human Genetics, 2016. 19(2): p. 87-96.

[13] de Manzano, Ö. and F. Ullén, Genetic and environmental influences on the phenotypic associations between intelligence, personality, and creative achievement in the arts and sciences. Intelligence, 2018. 69(p. 123-133.

[14] Zietsch, B.P., T.R. de Candia, and M.C. Keller, Evolutionary behavioral genetics. Current Opinion in Behavioral Sciences, 2015. 2(p. 73-80.

[15] Pryor, J.B., Sexual harassment proclivities in men. Sex Roles, 1987. 17(5): p. 269-290.

[16] Ménard, K., S., N.E. Shoss, and A.L. Pincus, Attachment and personality predicts engagement in sexual harassment by male and female college students. Violence and Victims, 2010. 25(6): p. 770-786.

[17] Stanford, M.S., et al., Fifty years of the Barratt Impulsiveness Scale: An update and review. Personality and Individual Differences, 2009. 47(5): p. 385-395.

[18] Lindsay, J.A., P. Boghossian, and H. Pluckrose (2018) Academic Grievance Studies and the Corruption of Scholarship. Areo Magazine, URL: https://areomagazine.com/2018/10/02/academic-grievance-studies-and-the-corruption-of-scholarship/ Date Accessed: 11 April 2019.

[19] Burrows-Taylor, E. OPINION: It’s not sexist to say this French MP shouldn’t wear revealing clothes on TV. 2018 [Last accessed: 1 April 2019]; Available from: https://www.thelocal.fr/20180227/opinion-its-not-sexist-to-say-mps-shouldnt-wear-revealing-clothes-on-tv.

[20] Bouchard, M. and P. Lussier, Estimating the Size of the Sexual Aggressor Population, in Sex Offenders: A Criminal Career Approach, A. Blokland and P. Lussier eds, 2015, John Wiley & Sons.

[21] Williams, W.M. and S.J. Ceci, National hiring experiments reveal 2:1 faculty preference for women on STEM tenure track. Proceedings of the National Academy of Sciences, 2015. 112(17): p. 5360-5365.

[22] Ceci, S.J., Women in Academic Science: Experimental Findings From Hiring Studies. Educational Psychologist, 2018. 53(1): p. 22-41.

[23] Ceci, S.J., et al., Women in Academic Science:A Changing Landscape. Psychological Science in the Public Interest, 2014. 15(3): p. 75-141.

[24] Ceci, S.J. and W.M. Williams, Sex Differences in Math-Intensive Fields. Current Directions in Psychological Science, 2010. 19(5): p. 275-279.

[25] Ceci, S.J. and W.M. Williams, Understanding current causes of women’s underrepresentation in science. Proceedings of the National Academy of Sciences, 2011. 108(8): p. 3157-3162.

[26] Ceci, S.J., W.M. Williams, and S.M. Barnett, Women’s underrepresentation in science: Sociocultural and biological considerations. Psychological Bulletin, 2009. 135(2): p. 218-261.

[27] Robertson, K.F., et al., Beyond the Threshold Hypothesis:Even Among the Gifted and Top Math/Science Graduate Students, Cognitive Abilities, Vocational Interests, and Lifestyle Preferences Matter for Career Choice, Performance, and Persistence. Current Directions in Psychological Science, 2010. 19(6): p. 346-351.

[28] Henrekson, M. and M. Stenkula, Why Are There So Few Female Top Executives in Egalitarian Welfare States? The Independent Review, 2009. 14(2): p. 239-270.

[29] Roll-Hansen, N., The Lysenko Effect: The Politics of Science. Control of Nature, ed. M.L. Schagrin, M. Ruse, and R. Hollinger. 2005, Amherst, NY: Humanity Books. 335.