Big things in small boxes

Two justifications I’ve seen for dismissing the relevance of computer ethics in decision making:

  • this decision is harmless,
  • this is a business decision not a technology one

Computer Ethics and Professional Responsibility
I’m reading a great collection called Computer Ethics and Professional Responsibility. In the essay Unique Ethical Problems in Information Technology Walter Maner writes:

In the discreet world of computing, there is no meaningful metric in which small change and small effects go hand in hand.” – (Dijkstra 1989. p. 1400) … the normally predictable linkage between acts and their effects is severely skewed by the infusion of computing technology.

This disconnect around human perception, size and complexity reminded me of John Maeda’s Laws of Simplicity:

Thus while [integrated circuits] are a primary driver of complexity in modern day objects, they also enable the ability to shrink a frighteningly complex machine to the size of a cute little gum-drop… There is no turning back to the age when large objects were complex and small objects were simple.

So in hard and soft technology, size is no longer a predictable measure of consequence. Small contains great expressive power.

From Bill Joy’s 2000 wired article on the threat of miniaturization, ubiquity and self-learning systems, Why the future doesn’t need us:

The cause of many such surprises seems clear: The systems involved are complex, involving interaction among and feedback between many parts. Any changes to such a system will cascade in ways that are difficult to predict; this is especially true when human actions are involved.

Not only small tools but mundane and seemingly insignificant decisions about their use can have tremendous consequences.

Should we limit our concern for consequence to harm: injury, economic loss and the like?

To think so would be to absolve most technologists of ethical considerations most of the time. But an ethical view point doesn’t limit itself to harm but considers net benefit. So this isn’t just a discussion of quality control, it’s one of maximizing value and not just to our employers, clients and ourselves but with consideration for all those who are affected by the use of our products.

But is good ethics good business?

If “good business” is maximizing return over a fixed period then that clearly is not the same as maximizing benefit to the larger society. Most would argue that there are meaningful considerations besides acquiring wealth for both an individual and an organization but I’ll leave that for now. Let me concede that in business there are winners and losers and that winners often gain at the expense of those around them.

Nonetheless, what Walter Maner, John Maeda and Bill Joy are telling us is that in a way unique to this time in history we cannot deem a thing or a thought innocuous because it is small.

To borrow from an ex-Secretary of Defense, our work is fraught with far more “known unknowns” and “unknown unknowns” and far less “known knowns” than our ape-evolved brains would have us believe.

Even out of narrow self interest we must approach decision making around technology with humility and not skip the hard work of thinking through the consequences of our actions.

  • this decision is harmless
  • this is a business decision not a technology one

Embrace such thoughts at your peril for you deny complexity that both envelops you and fits on the tip of your finger.

Women & Agile Development (revisit)

ACM Technews referenced this article from

IT still trying to find what women want

The world is flat. Better go out and hire some women.

That seems to be the gist of a recent Gartner report on the gender gap in information technology

The article lists five gender-based traits CIOs should pay attention to when building IT staffs.

There’s a danger of drawing conclusions about individuals from differences in large samples. Just because men are statistically greater risk takers than women doesn’t mean I’m bolder than any particular woman. Or that she has better listening skills.

But at the level of an entire industry statistical differences are meaningful. Software would be better if we had collaborated and communicated better. Our products would be better if we had more empathy for customers and end users. Attracting qualified women into the field is a contribution to that end.

As the article suggests, the only appropriate way to do that in the context of a specific hiring decision is to include people skills as a requirement of the position, cast a wide net and hire the best candidate.

But you need to have a workplace and compensation package that is attractive to someone with people skills, technical chops and wants to give you their best at a sustainable pace.

To the argument that inviting women into a dysfunctional IT culture won’t make things better, I’d say that’s not my point. More women in IT is a hoped for result of humane workplaces — not a solution for creating them.

We have to work from within to make our companies better. In my software development team, our use of Scrum and XP has helped us do that. So far, my employer has been receptive.

I know that isn’t true everywhere. All I can say is those of us who have a choice sometimes have to make tough decisions about where we choose to collect our paycheck.

Power, dissent, and bullying in software developer communities

Grassroots developer communities form around shared values in dissent against institutions and norms that dehumanize their work and diminish their efforts. They attack these orthodoxies with humor, heretical thinking, and hard work.

This benefits society when developers defy those with greater power. It harms society when developers bully people with less power.

At the ThoughtWorks sponsored Agile East, Martin Fowler spoke to his post, SmutOnRails.

Part of the community was offended by a presentation at the GoGaRuCo (Golden Gate Ruby Conference). Others fought back saying that no offense was meant, the presenter apologized, and that the tone was in the spirit of the Rails community.

(T)he view of the rails leadership seems to be this: that the objections to the presentation are yet another attempt to foist empty corporate values on the thriving Rails ecosystem… (more)

This debate is not unique to the Rails community. It reminds me of concerns my friend, Luke Melia, raised over jokes and behavior at the first Austin Alt.NET. Martin Fowler links off to a similar controversy in the Flash community.

It is also not unique to developer communities but developers in particular need to be concerned about the outcome.

Women, African Americans and Hispanics are under-represented in IT and even more so in software development. In 2001-2002 74.4% of software developers were men. 78% of those men were white.

In 1986 the percentage of women in CS programs peaked at 37%. The percentage of women in computer science programs has gone down since then.

In 2001-2, only 28 percent of all undergraduate degrees in computer science went to women. By 2004-5, the number had declined to only 22 percent. — What Has Driven Women Out of Computer Science?, NY Times

There were 15,000 women in CS progreams in 1986. Riding natural cycles this number was not matched again until 2003. This latter number contains a higher percentage of non-resident aliens who will not necessarily contribute to the US workforce.

This despite higher percentages and numbers of women acquiring college educations than men. In 2007, 33% of women 25-29 held a four year degree or higher versus 26% of men. 55% of graduates with four year degrees or higher aged 25-29 were women.

Women are even receiving the majority of degrees in science and technology. They have shown steady progress in biology, chemistry, physics, mathematics and engineering.

Metrics can be misinterpreted but these quantitative measures support a stunningly obvious anecdotal observation. US software developers are a white male enclave.

This is a power imbalance and we developers are part of the problem.

Isolation is a key factor for a higher attrition rate among women and minorities, said Teresa Dahlberg, director of the Diversity in Information Technology Institute at UNC Charlotte. People tend to associate with “like communities,” where people have similar backgrounds and interests, she explained. — Computer science lacks women, minorities, SD Times

So when we behave in a way that marginalizes and intimidates talented women and minorities, we abuse power. We become bullies. We are oppressors.

“There is a good amount of research that shows that women are judged more harshly than men, for hiring, evaluations and promotions,” she added. “Virginia Valian [author of "Why So Slow? The Advancement of Women"] shows this for women in science, technology, engineering and math faculty jobs.” Virginia Valian is a professor at Hunter College. — SD Times

Part of the problem may be a perception that software development doesn’t contribute enough to society. To the degree this perception is true it is damning. To the degree it is just a perception we have work to do as advocates.

Our actions need to be judged not by our intentions but by the outcome.

Requisite variety within our teams remains an essential enabling condition for sustained innovation.

Access to technology is growing across all tiers of class, race and gender both in the US and overseas. Diverse teams can better address our market and build software better adapted to our end users.

A more diverse workforce provides the kind of social change that will help us create a more humane workplace for developers.

Finally, anything that limits the number of able US software developers hurts our ability to compete.

When developer communities marginalize women and minorities, we conspire to isolate ourselves from the larger society. We defeat our own attempts to change the power structures around us and improve our lot and our output.

Are we driving women away from software development?

These are notes from my presentation at the Hawaii International Conference on System Sciences (HICSS) #45 on Agile values, product innovation and the shortage of women software developers.

I’ve broken the fifty slide, eighteen minute presentation into several posts.

This first part uses existing research to establish:

  • women are under-represented in software development,
  • this is a multi-decade trend atypical of Science, Technology, Engineering and Math (STEM),
  • women are leaving mid-career in disproportionate numbers and
  • young women are opting out as early as middle and high school.

I’ll link to my full paper when it is available and to subsequent posts as I publish them.

There is an abundance of research on the problems women face in our field. I would love real researchers to jump in on whether Agile principles and Agile practioners can really make a difference here.

I’d also love any suggestions of organizations, institutions and individuals I might reach out to for more information, collaboration, or to take up the cause.

(1-2) Hello,

My premise is the lack of women developers in the US is an impediment to value delivery and product innovation in the software industry.

In light of this, Agile principles call on practitioners to confront hostile workplace conditions and enterprises to address the material impediments of pay and advancement.

This beneficial change in teams and companies can incrementally change perceptions in the larger society.

(3) To introduce myself

I’m a software practitioner not a consultant or educator. I’ve studied and applied Agile methods for nine years.

I’ve spent most of my career in woman run organizations.

(4) I have a daughter

…who loves technology and sought out a culturally diverse, math and science school in Brooklyn. So, this topic is personal to me.

C. Hayes. “The Incredible Shrinking Woman” in Gender Codes : Why Women Are Leaving Computing. T. Misa, Ed. Hoboken, N.J. IEEE Computer Society, 2010, pg. 33

(5) Women are underrepresented in Computer Science

According to the US Bureau of Labor Statistics, women represent 46% of the workforce but only 25% of software developers. Over two decades the percentage of women developers has steadily declined.

S. Hewlett and C. Luce,. “The Athena Factor.” Harvard Business Review, pp. 51, Jun. 2008.

(6) Women are leaving mid-career

Women are leaving IT in larger numbers than men. 56% of women leave mid-career across all technology occupations. 41% leave their careers in “high technology” compared to only 17% of men. Half of women leaving STEM careers leave the STEM sector completely.

Four Decades of STEM Degrees, 1966-2004: ‘The Devil is in the Details.’” CPST, Sep. 2006, pg. 3.

(7) Women are not studying Computer Science

According to the Commission on Professionals in Science and Technology, in the decade between 1986 and 1995 the number of women earning Computer Science bachelor’s degrees dropped 55%. As of 2010, the percentage was still falling despite growing percentages of women graduating from four year colleges. This is not typical of STEM where 49% of bachelor degrees go to women.

N. Zarrett and O. Malanchuk. “Examining the Gender Gap in IT by Race” in Women and IT. J. Chohoon and W. Aspray Ed. Cambridge, MA. The MIT Press, 2006, pp.55-88.

(8) Young Women are disinterested in pursuing high tech education or careers

The Maryland Adolescent Development in Context Study, a longitudinal study of 1,400 white and african american students found that women were much more likely to have no interested in IT related careers and degrees than men.

(9) At what cost to the software industry?

In the last decade, the U.S. software industry represented $200B in annual sales and employed 2.2M software professionals. McKinsey & Co estimates in this decade, demand for mid-career IT professionals will increase by 25% while the available pool will decrease by 15%. This in a country where 71% of workers are in jobs with low demand or an oversupply of eligible candidates.

(10) The cost of attrition

And let’s also get at the cost of attrition. According to HR magazine, it costs approximately 100-125% of an employee’s annual salary to replace them. Retaining one-quarter of the women who leave computer engineering mid-career could represent a ten year savings of $8B to the industry.

Next: Are women are under-served by software…

The full citation list for my paper.

Are women are under-served by software?

These are notes from my presentation at the Hawaii International Conference on System Sciences (HICSS) #45.

I’ll link to my full paper when it is available and to subsequent posts as I publish them.

Agile values, product innovation and the shortage of women software developers Part 2 of 7

(11) Lost opportunity in the software industry (Product)

The lack of women on software teams is also a potential loss to product innovation…

(12) Women are our customer

Women directly or indirectly influence 61% of U.S. consumer electronics purchases[18].

(13) Women in gaming

Women are 42% of active game players and 48% of frequent game purchasers. And if you think they’re just buying them for their kids, industry research shows women 18 and over are 37% of game players whereas boys 17 and under are only 13%[19].

(14) Women on the internet

Half (50.4%) of the internet population is women 18 and over. They spend an average of 38 hours per month online. They spend 5% more time than men in online social networking and 20% more time on online shopping. Women account for 58% of internet buyers, 61% of internet transactions and 58% of internet dollars.

(15) Women are underserved

Software products are generally designed with no consideration for women as distinct user groups. In “Gender differences in Web Usability”, Frank Spillars states, “Gender differentiation is barely present in North American technology product design… let alone Web experiences[22].”

(16) how women perceive and use software

In “Towards Female Preferences in Design.” the authors found differences in the ways men and women perceive and describe software products. “The results of this research have revealed female-oriented themes that should… enlarge views of pleasurable product design attributes and language for the genders[23].”

(17) three ways companies fail.

Boston Consulting Group (BCG) highlights three ways companies fail to address women consumers: Poor product design: failing to tailor products to women’s unique needs and challenges.

(18) three ways companies fail.

Clumsy sales and marketing: based on outdated images and stereotypes.

(19) three ways companies fail.

Inability to provide meaningful hooks or differentiation: considering women indistinguishable from the general customer population or thinking of them as one monolithic segment[24].

Next: Can women devs help software better address the needs of women end users…

Previous: Are we driving women away from software development?

All slides published to date.

There is abundant research on the problems women face in our field. I would love researchers to jump in on whether Agile principles and Agile practioners can really make a difference.

I’d also love any suggestions of organizations, institutions and individuals I might reach out to for more information, collaboration, or to take up the cause.

Please comment on my proposal to Agile 2012.

The full citation list for my paper.