Decision-Making Revisited

Welcome back. More than a decade ago, I blogged about my decision to leave academia for the federal government. I wrote, How are you at making decisions? I’m big on intuition (ask Myers or her mother, Briggs); but I’m kind of wishy-washy if I don’t have a mountain of information and the results of surveying at least half the people on the East Coast. (Decision-Making Time).

A recent study by researchers affiliated with Stevens Institute of Technology and Lehigh University suggests my decision-making needs refinement. They show that most people’s decision-making actually gets worse, not better, when they’re given additional facts and details.

Modeling Decision-Making
To study how people make decisions, researchers typically create causal models--diagrams that show how different factors interact to yield specific outcomes. For these, the modelers choose either the fullest accounting of the causal structure, which may overwhelm decision makers, or simplified, targeted versions that omit miscellaneous detail.

Example of a complex causal model diagram with all the bells and whistles for managing a healthy body weight (Fig. 1 from cognitiveresearchjournal.springeropen.com/articles/10.1186/s41235-023-00509-7).

Studying People’s Decision-Making
For the recent study, the researchers conducted a series of experiments to explore how people’s decision-making varies when they’re presented different kinds of causal models of real-life topics (e.g., buying a house, managing body weight). The experiments enlisted from 300 to 800 US adults from the online Prolific survey platform before screening.

They first tested how the level of detail in a causal model influences decision accuracy. With no model (no diagram) as the control, they compared models targeted specifically to each decision to more complex models. Although targeted models might confuse if information that participants expect to see is omitted, the more complete models are lacking if participants can’t determine what part of a model (what part of the diagram) to use or if they’re unable to ignore irrelevant information.

 Targeted causal model for donating to charity (Fig. 4 from cognitiveresearchjournal.springeropen.com/articles/10.1186/s41235-023-00509-7).

The researchers then tested whether directing people’s attention to the relevant paths within more complex models--highlighting the paths--would gain the benefits of targeted information.

Complete causal model for donating to charity highlighted with the targeted path (Fig. 4 from cognitiveresearchjournal.springeropen.com/articles/10.1186/s41235-023-00509-7).

Finding that providing targeted or highlighted-path diagrams that focus on the information needed for a decision can lead to better choices than either complex diagrams or no diagrams, they tested the boundaries of this effect. They found that adding even a small amount of information beyond that related to the targeted answer was detrimental.

Wrap Up
Well, I’m impressed with causal models. What’s more important, I understand that overloading decision-making with information is not the way to go. But I’m more than a bit shaky about developing the targeted causal model to begin with.

As I wrote in that long-ago blog post, Lining up the pros and cons with my wife, who is rarely decision-challenged, didn’t produce any clear answer. One never knows all the pros and cons, but in some cases, we weren’t even sure which was which.

Thanks for stopping by.

P.S.
Study of causal models in Cognitive Research: Principles and Implications: cognitiveresearchjournal.springeropen.com/articles/10.1186/s41235-023-00509-7
Article on study on EurekAlert! website: www.eurekalert.org/news-releases/1002995

Intelligence vs Thinking Speed

Welcome back. Today’s question: Do people with higher IQs think faster than those with lower IQs?

Sorry, that’s sort of a trick question from the results of a recently published study. The study was part of an effort to simulate the human brain using computers to understand how decision-making processes work and why different people make different decisions.

The researchers were affiliated with the Brain Simulation Section at the Berlin Institute of Health at Charité and the Department of Neurology and Experimental Neurology of Charité--University Hospital in Berlin, with a colleague at the Center for Brain and Cognition, Computational Neuroscience Group, University of Pompeu Fabra in Barcelona.

Brain simulation helps us understand how our brain works (from brainsimulation.charite.de/en/).

Personalized Brain Models
The Berlin researchers begin with a human brain model derived from digital brain scan data, such as magnetic resonance imaging (MRI), and from models based on theoretical knowledge of biological processes. The next step is to refine the general model using data from individuals to produce “personalized” brain models.

Personalized brain models reproduce the activity of the participants’ brains very effectively. They behave differently from one another in the same way as their human counterparts, matching intellectual performance and reaction times.

For the current study, the researchers used data from 650 participants of the Human Connectome Project, a U.S. initiative that has been studying the brain’s neural connections for over a decade. The participants had undergone extensive cognitive testing, and their IQ scores were known.

Higher vs Lower IQ Thinking Speed
In one test (Penn Matrix Reasoning Test), the participants were asked to identify logical rules in a series of patterns. These rules became increasingly complex with each task and more difficult to decipher.

(Answer to trick question) Participants with higher intelligence scores solved simple tasks faster than people with lower IQ scores, but they took longer to solve difficult tasks.

Thinking Speed and Brain Synchronization
MRI scans showed that “slower” solvers’ brains were more synchronized; they had higher average functional connectivity between brain regions. This altered the properties of working memory and thus the ability to endure prolonged periods without a decision.

In essence, tackling more challenging tasks, higher IQ, slower solver brains store progress in working memory while they explore other solutions before integrating all to arrive at a decision. It may take longer yet the results are better.

In contrast, the personalized brain simulations allowed the researchers to determine that brains with reduced functional connectivity literally “jump to conclusions.” Lower IQ brains make decisions, wrong or right, without waiting for upstream brain regions to complete the processing steps needed to solve the problem.

Wrap Up
How are you at decision making?

Me? As I blogged in 2012, I’m big on intuition (ask Myers or her mother, Briggs); but I’m kind of wishy-washy if I don’t have a mountain of information and the results of surveying at least half the people on the East Coast (see: Decision-Making Time).

Thanks for stopping by. Do decide to visit again.

P.S.
Study of how brain network shapes decision-making in Nature Communications journal: www.nature.com/articles/s41467-023-38626-y
Article on study on EurekAlert! website: www.eurekalert.org/news-releases/991304
Human Connectome Project and Connectome Coordination Facility: www.humanconnectome.org/about-ccf

End-of-Life Caring

Comfort care is an essential part of medical care at the end of life…The goals are to prevent or relieve suffering as much as possible and to improve quality of life while respecting the dying person's wishes. (National Institute on Aging)
 

End-of-life care for the terminally ill
(photo from www.nia.nih.gov/).

Welcome back. Although I’m not yet dying or attending to someone who is, I came across two recent end-of-life studies you might find of interest. One focuses on the dying patient and family, the other addresses the surrogate who makes life-sustaining decisions for the patient.

The 3 Wishes Project
The 3 Wishes Project is an end-of-life program that seeks to bring peace to terminally ill patients and ease the grieving process.

The program involves implementing wishes identified by the patient, family, clinicians or project team in an effort to dignify the death and celebrate the life; humanize the dying process and create positive memories; and foster patient and family-centered end-of-life care while inspiring a deeper sense of vocation for clinicians.

The 3 Wishes Project began at St. Joseph's Healthcare Hamilton, an academic and research hospital affiliated with McMaster University and Mohawk College, in Hamilton, Ontario, Canada.

Can Project Sites be Added?
Researchers led by those with McMaster University conducted a study to determine if the program could be implemented by intensive care units of other hospitals.

Three additional hospitals participated, one each in Toronto, Vancouver and Los Angeles. Together with the Hamilton hospital, the study fulfilled 3,407 wishes for 730 dying patients.

The wishes, usually more than three per patient, included bringing personal items, pictures and pets from home, providing favorite music or spiritual support, connecting long lost family, celebrating weddings, watching a sporting event together with a favorite beverage, and a "date night" with local restaurant food.

3 Wishes Project--family, friends and staff get together in patient's room (photo from brighterworld.mcmaster.ca/articles/project-to-answer-last-wishes-spreads-successfully/).

Judging Success
The researchers assessed results using a mixed-methods formative evaluation, which entailed collecting, analyzing and integrating quantitative and qualitative data from 75 family members, 72 clinicians and 20 managers or hospital administrators.

Program value encompassed comforting families while inspiring compassionate clinical care. Transferability was promoted by family appreciation and the intensive care unit culture committed to dignity-conserving, end-of-life care. As for affordability, there was a required minimal investment for reusable materials, but the average cost per wish was just over $5.00 since most wishes cost nothing. Sustainability was demonstrated by each site continuing the program after the study.

Clinician and family perspectives on the 3 Wishes Project
(photo from 11-minute video youtu.be/CkWjlcl4BA4).

There seems no question that the 3 Wishes Project can and should be implemented at other tertiary care centers.

Religion, Spirituality and Surrogate Decisions
A team of investigators, led by a researcher with the Regenstrief Institute, set out to determine the relationships between religion and spirituality and the treatment decisions made by health care surrogates.

Decision-makers consent for Do-Not Resuscitate status
(from Patricia Bomba’s slides on “Medical decision-making
capacity: Legal, Ethical and Clinical Considerations”
slideplayer.com/slide/4174065/).

They enlisted 291 patients and their health care surrogates from three hospitals. The patients were age 65 or older and admitted to the intensive care services. The surrogates were predominately Protestant.

Baseline surveys completed between the second and tenth day assessed dimensions of religion and spirituality. Review of medical records and health information six months later identified life-sustaining treatments and hospice for patients who died.

Key Factors Influencing Surrogates
After adjusting for other religious dimensions, demographic and illness factors, the surrogates' belief in miracles was the only factor significantly associated with lower preference for do-not-resuscitate status--59% believed a miracle might save the patient.

Higher surrogate intrinsic religiosity (religion that is an end in itself) was associated with lower receipt of life-sustaining treatments during the patients’ final 30 days.

Together, belief in miracles and higher intrinsic religiosity were associated with lower hospice utilization.

To reduce effects on end-of-life treatment, the investigators recommend that chaplains or appropriately trained clinicians identify and explore surrogates’ belief in miracles and intrinsic religiosity.

Wrap Up
If you’re seeking end-of-life information, you’ll find an excellent series of articles on the National Institute on Aging’s website--providing care and comfort, palliative and hospice care, caring for a dying relative or someone with dementia, healthcare decisions, what happens when someone dies, what to do after someone dies and mourning the death of a spouse.

Thanks for stopping by.

P.S.
National Institute on Aging’s End of Life website: www.nia.nih.gov/health/end-of-life
3 Wishes Project website: 3wishesproject.com/
3 Wishes Project study in Annals of Internal Medicine: annals.org/aim/article-abstract/2755629/compassionate-end-life-care-mixed-methods-multisite-evaluation-3-wishes
Article on study on EurekAlert! website: www.eurekalert.org/pub_releases/2019-11/mu-pta110619.php
Surrogate decision maker study in Journal of Pain and Symptom Management: www.sciencedirect.com/science/article/abs/pii/S0885392419305263
Article on study on EurekAlert! website: www.eurekalert.org/pub_releases/2019-11/ri-fso110419.php

Driverless Car Ethics

Welcome back. When my son, Noah, was working his way through a B.S. in Mechanical Engineering, we seldom spoke about his technical courses. But I do remember discussing one assignment he had in an elective humanities course; it pertained to what I now know to be labeled the trolley problem.

Although this thought experiment in ethics has been around for over a century with any number of variants, consider the following: A runaway trolley is going to hit several people. You can pull the lever, directing the trolley to a sidetrack, where it will strike only one person. Should you pull the lever? Is there a moral difference between doing harm and allowing harm to happen?

The basic trolley problem: Do nothing and trolley hits five people, or pull lever to redirect trolley to a sidetrack, where it will hit one person (from nymag.com/intelligencer/2016/08/trolley-problem-meme-tumblr-philosophy.html).

Noah graduated and moved on; however, the trolley problem is now being addressed in different studies of programming driverless vehicles (aka self-driving or autonomous vehicles). I thought you’d find examples of that work of interest.

The Social Dilemma
A 2016 study by researchers from France’s University of Toulouse Capitole, the University of Oregon and MIT examined the trolley problem in six online surveys of 182 to 393 U.S. participants.

Overall, the participants seemed to agree that autonomous vehicles (AVs) should be programmed to be utilitarian, minimizing the number of casualties. Yet given the incentive for self-protection, few would be willing to ride in utilitarian AVs. Further, they would not approve of regulations mandating self-sacrifice, and such regulations would make them less willing to purchase an AV.

German Guidelines

Report of German Ethics
Commission on Automated
and Connected Driving
(see link in P.S.).

Trying to stay ahead of the issue, Germany’s Federal Minister of Transport and Digital Infrastructure appointed an Ethics Commission on Automated and Connected Driving.

The commission’s 2017 report included 20 ethical rules. One, for example, states in part: In the event of unavoidable accident situations, any distinction based on personal features (age, gender, physical or mental constitution) is strictly prohibited…

Study Results at Odds with Guidelines
A 2018 study by researchers from Germany’s Osnabrück University found that what is morally justified may not be socially acceptable. Their study had 189 participants (average age 24) complete several virtual reality simulations of driving alone on different two-lane roads. Obstacles emerged on both lanes giving the participants four seconds to switch or not switch lanes before hitting someone (which wasn’t shown).

The study found nearly all participants would change lanes to hit fewer people; over half would sacrifice themselves to save others, especially as the number saved grew; most would hit an elderly person before an adult, more so before a child; and most would swerve onto the sidewalk to save a greater number of people.

Introducing Probabilities
A more recent study by a research team from Germany’s Max Planck Institute for Human Development and University of Göttingen examined trolley problem options when the probabilities of hitting the pedestrian or bystander were known or unknown (872 U.S. participants). They also considered how people retrospectively evaluate those options when a road user has been harmed (766 U.S. participants).

They found that participants placed particular weight on staying in the lane. This tendency was seen when probabilities were known or uncertain and in hindsight after accidents occurred. Staying in the lane was considered more morally acceptable, particularly for autonomous vehicles.

International Survey
In the most recently published study, collaborators from MIT, Harvard, the University of British Columbia and University of Toulouse Capitole reported nearly 40 million trolley-problem decisions made by people from 233 countries.

MIT’s Moral Machine, an online experimental platform, presented the trolley problem in ten languages, examining nine scenarios (e.g., saving more vs. fewer lives).

Example Moral Machine question: Driverless car’s brakes fail; car will kill 3 elderly people (note skulls!) crossing on a do-not-cross signal (left). Swerving will hit barrier and kill 3 passengers--2 adults and 1 child (from www.nature.com/articles/s41586-018-0637-6).

Two key findings were:
- Globally, the strongest preferences are for saving humans over animals, more lives and young lives.
- Three distinct moral clusters of countries could be identified, suggesting that groups of territories might converge on shared preferences, while between-cluster differences may pose problems.

Wrap Up
Research on programming driverless cars continues. The MIT-led study noted they could not do justice to the complexity of autonomous vehicle dilemmas, even with the large sample they obtained--we need to have a global conversation to express our preferences to those who will program the vehicles and to those who will regulate them.

Thanks for stopping by.

P.S.
Trolley Problem:
www.wired.com/story/self-driving-cars-will-kill-people-who-decides-who-dies/
en.wikipedia.org/wiki/Trolley_problem
2016 University of Toulouse Capitole-led study in Science and article on Quartz website:
science.sciencemag.org/content/352/6293/1573
qz.com/536738/should-driverless-cars-kill-their-own-passengers-to-save-a-pedestrian/
German Ethics Commission on Automated and Connected Driving report:
www.bmvi.de/SharedDocs/EN/publications/report-ethics-commission.pdf?__blob=publicationFile
2018 Osnabrück University study in Frontiers in Behavioral Neuroscience and article on ScienceDaily website:
www.frontiersin.org/articles/10.3389/fnbeh.2018.00031/full
www.sciencedaily.com/releases/2018/05/180503142637.htm
2018 study with probabilities added in Risk Analysis and article on ScienceDaily website:
onlinelibrary.wiley.com/doi/abs/10.1111/risa.13178
www.sciencedaily.com/releases/2018/10/181009135828.htm
2018 Moral Machine experiment in Nature and articles on study:
www.nature.com/articles/s41586-018-0637-6
www.sciencedaily.com/releases/2018/10/181024131501.htm
www.technologyreview.com/s/612341/a-global-ethics-study-aims-to-help-ai-solve-the-self-driving-trolley-problem/
Moral Machine (go ahead and try it): moralmachine.mit.edu/

A version of this blog post appeared earlier on www.warrensnotice.com.

Waiting for Income

Save your money. (Barney Large Coin
Piggy Bank from www.amazon.co.uk)

Welcome back. Today’s question is Why do some people make more money? If I may get personal, what factors had the greatest effect on your earnings? Go ahead, fill in the blanks.

Researchers from Temple University did. Along with the expected results, such as education and occupation, they found one interesting surprise: Delay discounting ranked high as a predictor of future income.

What’s Delay Discounting?
If you had to choose between receiving $1 today or $2 tomorrow, would you take the $1? Those who take the smaller reward today rather than wait for the larger reward tomorrow are discounting the value of the future reward. Delay discounting refers to how much a person devalues future rewards compared to present rewards.

The classic studies of delayed gratification--the marshmallow experiment--were conducted at Stanford University in the late 1960s and early 1970s. Children were offered one reward (e.g., a marshmallow) immediately or two rewards if they waited about 15 minutes while the person administering the test left the room. Get this: Follow-up studies found children who waited longer tended to have better life outcomes.

Factors Related to Future Income
To improve upon earlier investigations of income attainment, the Temple researchers tested a large, diverse population: 2,564 racially and ethnically heterogeneous, male and female participants, 25 to 65 years in age, pre-high school to PhD in education, $10,000 to $235,000 in annual income and from over 1,700 zip codes.

In addition, they employed a novel analytic approach, using three machine-learning algorithms to model the relationship between income and key factors identified in earlier research--age, gender, ethnicity, height, race, zip code, education, occupation and delay discounting behavior.

Measuring Delay Discounting
Test participants were initially asked to choose between $500 immediately versus $1,000 at five different delays (1 day, 1 week, 1 month, 6 months, 1 year). If they chose the immediate reward, the next question offered an immediate reward midway between the prior immediate reward and zero. If they chose the delayed reward, the next question offered an immediate reward midway between the prior immediate reward and $1,000.

This narrowing pattern continued until participants’ choices converged on the dollar amount subjectively equivalent to the discounted delayed reward if the value were offered immediately. Lower dollar amounts indicated increased devaluation of delayed rewards in favor of immediate rewards.

Study Findings
Modeling with the three machine-learning algorithms, the researchers found that individual differences in income were explained by factors that could be ranked in a consistent manner.

Average ranking of factors according to how well they predicted salary by three machine-learning algorithms (from www.frontiersin.org/articles/10.3389/fpsyg.2018.01545/full).

Occupation and education were paramount with each algorithm, and on average, zip code and gender were the next most important factors. The fifth most important factor was delay discounting, which was more predictive than ethnicity, height age and race.

One study shortcoming is that representation of African Americans and Hispanics in the sample population was only about half that in the U.S. population at large.

Wrap Up
As to why individual differences in discounting of future rewards predicts income attainment, the researchers speculate that it may be a consequence of the correlation between higher discounting and undesirable life choices.

Difficulties delaying gratification may also be affected by episodic future thinking, i.e., the ability to project oneself into the future to pre-experience an event. If people can vividly imagine themselves in the future with the larger rewards, they are more likely to be patient.

So, relax. As you’ve likely heard, all things come to those who wait. Thanks for stopping by.

P.S.
Study of delay discounting for income attainment in Frontiers in Psychology journal: www.frontiersin.org/articles/10.3389/fpsyg.2018.01545/full
Articles on study on ScienceDaily and ScienceAlert websites: www.sciencedaily.com/releases/2018/09/180903101741.htm
www.sciencealert.com/your-ability-delay-instant-gratification-predict-money-earn-delay-discounting-marshmallow-test
Marshmallow experiment: en.wikipedia.org/wiki/Stanford_marshmallow_experiment
Update study of children’s delay of gratification in Developmental Psychology journal: psycnet.apa.org/record/2018-29923-001?doi=1

A version of this blog post appeared earlier on www.warrensnotice.com.

Who Chooses to Lead?

Welcome back. Early in my years in government, I had a mandatory course on leadership styles and behavior. It was one of the few courses I enjoyed, especially compared to the many I had on contracts and project management.

Along with learning to give feedback, which I never quite mastered, and being exposed to a raft of other long-forgotten tips, my leadership style was characterized as “coaching.” That was hardly surprising since I’d spent the previous 20 years in academia teaching.

I thought of that course, not the style or behavior of our country’s leader, when I saw a study on leadership. Researchers from Switzerland’s University of Zürich and ETH Zürich examined what determines someone’s willingness to lead, to take responsibility for others.

Who takes the lead? (photo from
www.aleanjourney.com/2013/09/lean-leadership-lessons-we-can-learn.html)

Testing Decision Preferences
The researchers collected a variety of behavioral, computational and neurobiological information from test participants.

For a baseline task, 40 participants completed 200 trials selecting either a risky or a safe gamble based on the probabilities of success and the possible gains or losses. In 140 of the trials, only partial information regarding the probabilities was provided.

Placed in groups of four for a delegation task, each participant was later presented the same choices from the 140 trials in both self trials and group trials (280 trials).

For these trials, participants had the option of deferring the selection of the risky or safe gamble to their group, which collectively had more information regarding probabilities of success. In the self trials, only the participant’s payoff was at stake; in the group trials, the selection determined the payoff of every group member equally.

One example of the 140 group trials of the Delegation Task, where the decision to Act could be deferred to the group. The probabilities of success (e.g., 20%) and failure (e.g., 30%) were shown by the proportion of green or red slices. In each trial, a varying amount of probability information was obscured (e.g., 50%), and different group members saw different information on the underlying probabilities. Choosing to Not Act gave an outcome of 0 for that trial. The same example appeared in a Baseline Task trial, where the decision could not be deferred. (figure from science.sciencemag.org/content/361/6401/eaat0036)

The researchers also collected functional magnetic resonance imaging (fMRI) data on a separate sample of 44 participants as they performed the same decision-preference tasks.

Decision Preferences and Leadership Scores
In addition to the decision-preference tasks, participants completed questionnaires on two widely used scales that predict leadership positions.

The researchers found no consistent correlation between the decision-preference measures and leadership scores. Preferences for decisions and control over self or others did not explain individual differences in leadership scores, suggesting the study should seek different motivational forces. They focused on a critical difference between self and group trials: the responsibility for others’ welfare.

Leadership and Responsibility Aversion
Deferrals, or what the researchers termed responsibility aversion, increased an average of 17% from self to group trials. That varied substantially across individuals (standard deviation of 43%). Most important, participants who deferred less--who showed lower responsibility aversion--had higher leadership scores.

To validate the association between responsibility aversion and leadership, the researchers collected factual expressions of the participants’ leadership behavior (mandatory military service rank, leadership experience in scouts’ organizations).

Responsibility aversion was the only measure that significantly correlated with these expressions of leadership.

Computational modeling of the decision-preference group data and functional connectivity modeling of the fMRI group data confirmed that responsibility aversion is a key determinant of the willingness to lead.

They postulate that responsibility aversion is driven by a cognitive process that reflects an increase in the demand for certainty about the best choice when others’ welfare is affected.

Wrap Up

High-scoring leaders. Five
presidents in the Oval Office
of the White House,
7 Jan 2009. (photo from CNN)

The researchers conclude that high-scoring leaders’ preferences for risk, loss and control can vary substantially. The unifying element is that the leaders calibrate their beliefs about decision-making similarly across self and group trials.

An authoritarian leader (think Trump) with a strong preference for control would have a very narrow deferral threshold in both self and group trials; an egalitarian leader (think Obama) with a strong preference for consensus would have a rather broad deferral threshold in both self and group trials.

The study’s findings provide a conceptual framework for an individual’s decision to lead--to assume responsibility for others’ outcomes--as well as insights into the cognitive and neural mechanisms driving the choice.

Thanks for stopping by.

P.S.
Study of leadership and responsibility in Science: science.sciencemag.org/content/361/6401/eaat0036
Article on study on ScienceDaily website: www.sciencedaily.com/releases/2018/08/180803103257.htm

A version of this blog post appeared earlier on www.warrensnotice.com.

Startup Funding Gender Gap

Congratulations! Your business is doing great! You tapped savings, family and friends to get started. You worked hard, recovered from early mistakes and found your market. Now, it’s time to go big, really big; you’re ready to go after venture capital. Oh, wait. I’m sorry. You’re a woman.

The hands for startup funding
sure look like those of men
(from corporatemonks.com/).

Welcome back. Let’s start in Europe. Sweden ranks at the top of the European Union’s Gender Equality Index, yet female-owned businesses, which account for one-third of Swedish businesses, receive only 7% of government venture capital.

A study by researchers affiliated with Sweden’s Luleå University of Technology and Halmstad University and Finland’s Hanken School of Economics documented how gender bias entered into the assessment of venture capital applications.

Venture Capitalist Stereotypical Notions
The study used interview data to first examine how 11 venture capitalists from two government organizations used notions of gender in assessing applications from 126 entrepreneurs (72 male, 54 female).

The researchers identified four gender-stereotypical notions:
-Women are cautious and risk-averse; men are ambitious and risk-taking.
-Women are reluctant to grow their businesses; men are willing to do so.
-Women do not have resources to engage in high growth; men do.
-Women’s ventures underperform; men’s ventures perform well.

To test these notions against fact, they statistically analyzed relevant performance indicators and accounting information for each of the 126 ventures. They found no significant difference between ventures led by men or women in any of the four identified areas: risk-taking, growth, growth resources or underperformance. In short, the perceived gender differences affecting funding decisions are myths.

OK, that’s what happens in Sweden, not the U.S. Wrong. A 2017 article in Fortune magazine reported that, of the billions venture capitalists invested in 2016, only about 2% went to women. That doesn’t appear to have changed much in 2017 (see figure).

Yearly venture capital funding ($ billions) for U.S. startups founded by men, women and both men and women, 2006-2017 (from fortune.com/2018/01/31/female-founders-venture-capital-2017/).

Male vs. Female Investors
Would male and female venture capitalists make different funding decisions? Maybe.

Research collaborators from the California Institute of Technology and University of California, San Diego, studied angel investors--individuals who invest their personal funds in a business; venture capitalists invest other people’s money.

Their analysis of a proprietary dataset from AngelList, a U.S. website for startups, angel investors and job-seekers, found male investors expressed less interest in female-led startups than in similar male-led startups. In contrast, the same female-led startups were more successful than male-led startups with female investors. The results did not appear to be driven by differences in startup quality, sector focus or risk.

So, adding female investors might help balance the gender funding gap, but there’s a long way to go. The Wall Street Journal highlighted a 2017 analysis of 71 top venture-capital firms that found less than 10% of their investment-team members were women.

Amateur vs. Professional Investors
An interesting sidelight to professional investing in startups is crowdfunding, where many small amounts of money are raised from a large number of amateur investors.

PwC in collaboration with The Crowdfunding Center analyzed 2015 and 2016 data from nine of the largest crowdfunding global platforms. They found that, although men sought startup crowdfunding more than women, women were more successful at reaching their funding target across a wide range of sectors, geography and cultures.

For example, considering total global campaign activity, men initiated about 2.5 times the number of startup campaigns as women, yet 22% of women-led campaigns were successfully funded compared to 17% of men-led campaigns. Data for the U.S. showed men initiated 2.3 times the number campaigns as women, yet 4% more women-led campaigns were successfully funded.

Total global crowdfunding activity in 2015 and 2016 for male- and female-led startups (from www.pwc.com/gx/en/about/diversity/womenunbound.html).

After finding similar results in their analysis of 416 projects from the crowdfunding platform Kickstarter, researchers from Louisiana State, Indiana and Suffolk universities conducted an experiment with 73 amateur investors to get a sense of why women were more successful than men. Female entrepreneurs were seen as more trustworthy, trustworthiness fostered funders’ backing and the funders’ implicit gender bias strengthened those effects.

Wrap Up
Getting back to professional investors, I’ll close with one bright note: Founders for Change. This is a loose, growing coalition of over 900 founders and chief executives who are improving diversity and inclusion within their companies and pressuring the venture capital industry to diversify.

Thanks for stopping by.

P.S.
Study of Swedish venture capitalists’ gender bias in Journal of Business Venturing Insights: www.sciencedirect.com/science/article/pii/S2352673417300938
Article on Swedish study in Harvard Business Review: hbr.org/2018/03/vc-stereotypes-about-men-and-women-arent-supported-by-performance-data
Fortune magazine article on venture capital funding gender gap: fortune.com/2017/03/13/female-founders-venture-capital/
Study of angel investors’ gender bias on Social Science Research Network website: papers.ssrn.com/sol3/papers.cfm?abstract_id=2953011
Wall Street Journal article on venture capital firm diversity: graphics.wsj.com/table/VCLEDER0410
PwC and The Crowdfunding Center report, Women unbound: www.pwc.com/gx/en/about/diversity/womenunbound.html
Study of crowdfunders’ trust in women in Journal of Business Venturing: www.sciencedirect.com/science/article/abs/pii/S0883902616302798
Article on crowdfunder trust study on ScienceDaily website: www.sciencedaily.com/releases/2018/05/180510101310.htm
Founders for Change: www.foundersforchange.org
New York Times article on Founders for Change: www.nytimes.com/2018/03/20/technology/founders-for-change-tech-diversity.html
Example articles on finding angel and venture funding:
articles.bplans.com/5-essentials-for-angel-investment/
articles.bplans.com/10-tips-finding-venture-funding/

A version of this blog post appeared earlier on www.warrensnotice.com.

Problem-Solving Squirrels

Vicki’s home-built,
squirrel-proof
bird feeder.

Welcome back. In one of my earliest blog posts (Lawn, Garden & Squirrels Photo Addendum), which was an addendum to a post voted exceptionally humorous (Time for Lawn and Garden), I described my wife Vicki’s struggles to defend birdseed from squirrels.

After failing with commercially available, squirrel-proof bird feeders, she painted and stood a pvc tube, about 8-foot tall, 8-inch diameter, on our deck and hung bird feeders and faux vegetation from its top.

As I wrote: Was it fear of jumping, missing and falling that restrained the squirrels, or were they just laughing too hard? Whatever it was, it worked for a couple of days.
 

Oops. A squirrel
made the leap.

Although I needed no further insight concerning squirrels’ problem-solving ability, researchers from the U.K.’s University of Exeter felt differently. They set out to examine how memory, together with behavioral traits, enhance squirrels’ problem-solving efficiency.

Testing Squirrels’ Memory
The researchers assessed five gray squirrels’ ability to retrieve food from the same puzzle box the squirrels had overcome 22 months earlier as well as from a physically dissimilar puzzle box that required the same actions for success.

The original puzzle box was a transparent plexiglass cube, outer dimensions about 10 inches. Ten holes were aligned vertically on each side, and 10 levers were inserted through holes on opposite sides. One end of each lever had a container for a hazelnut positioned just inside the box; 5 levers held a nut, 5 were empty. The box stood on legs, allowing space for squirrels to obtain nuts that fell from the containers.
 

Squirrel-testing puzzle boxes. Top: original box, first used 22 months earlier; Bottom: two views of new box. (Photo from link.springer.com/article/10.1007%2Fs10071-017-1113-7)

The most effective squirrel behaviors to obtain a nut were pushing the end of a lever near the container or pulling the opposite end. Pulling the near end or pushing the far end wouldn’t work.
 
The new puzzle box added for the study was an A-frame, four-sided triangular prism. There were only 5 levers inserted through side holes, which were horizontally, not vertically aligned. The effective and ineffective squirrel behaviors for retrieving a nut still applied, though all of the levers had nuts.

The squirrels (Arnold, Leonard, Sarah, Simon and Suzy) were lab residents, average age 6 years, with similar experimental histories in cognitive tasks. The researchers had them participate individually in a series of trials that began when the squirrel touched the box; it ended when the squirrel obtained all the nuts or specified times had elapsed.

Squirrel Performance
Encountering a new stimulus, the squirrels took on the order of 23 seconds to make contact in the first trial with the new puzzle box, compared to about 11 seconds in the last trial with the original puzzle box.

Once they got started, however, the squirrels needed about half the time to retrieve nuts in the first trial with the new puzzle box than they did in the first trial with the original box (2 sec versus 4 sec) and only about 1 second more than they needed in the last trial with the original box.

The researchers used video to analyze four behavioral traits: persistence (rate of attempts), selectivity (proportion of effective behaviors), motor diversity (rate of trying different tactics) and flexibility (rate of switching tactics after a failed attempt).

All squirrels demonstrated a high proportion of effective behaviors, reflecting the interaction between memory and behavioral traits for problem-solving. Remembering task-effective tactics, they consistently changed from ineffective to effective behaviors after failed attempts.

Wrap Up
Now that the role of memory in the problem-solving ability of squirrels (at least five) has been addressed, it’s appropriate to ask: Do squirrels find the nuts they hoard?

Here, the findings are mixed. I’ve seen comments about studies that show squirrels fail to recover most of the nuts they bury, but I haven’t seen those studies. In contrast, small sample, controlled studies I’ve reviewed had the opposite results. One article posited that just because a squirrel hasn’t retrieved a nut doesn’t mean that it won’t.

Perhaps the answer is muddied by the generally agreed findings that squirrels repeatedly rebury nuts to deter thieves and pretend to bury a nut to deceive onlookers. They’re pretty smart. Thanks for stopping by.

P.S.
University of Exeter study in Animal Cognition journal: link.springer.com/article/10.1007%2Fs10071-017-1113-7
Article on study on ScienceDaily website: www.sciencedaily.com/releases/2017/07/170713154843.htm
Example studies and articles on squirrels’ ability to locate hoarded nuts:
- www.sciencedirect.com/science/article/pii/S0003347205805068?via%3Dihub
- www.sciencedirect.com/science/article/pii/S000334729690528X?via%3Dihub
- animals.mom.me/smart-squirrel-6321.html
- www.nytimes.com/1994/12/11/nyregion/cuttings-now-it-can-be-told-all-about-squirrels-and-nuts.html
- en.wikipedia.org/wiki/Eastern_gray_squirrel

Drug Safety

Welcome back. Maybe it’s me and my TV viewing habits. Other than Netflix offerings in the evening, I watch nothing but news, though I am eclectic about the network--Bloomberg, CNN, MSNBC, Fox, NBC et al.--even if I often mute the sound and just read the news tapes.

Whatever the reason, I’m seeing an increasing number of ads for drugs, particularly prescription drugs. Presumably, TV viewers are supposed to see the ad, pay attention if the drug treats a malady they or a loved one have, remember the product name and suggest it to their physicians, because the billions of dollars pharmaceutical companies spend marketing directly to physicians isn’t enough.

New drugs are continually coming
to market. (multiple websites)

Given advertising costs and the need to promote new products, it’s likely that what’s being pitched via my TV is in some way new, if not seasonal.

If the new drug is intended to replace an older drug for my malady, my physician and I have to ask, Should I switch? The answer implicit in the results of a study published last May is, Maybe not.

The study was conducted by a team of medical researchers led by an investigator from Brigham and Women’s Hospital in Boston. And once again, I’m indebted to the University of California Berkeley Wellness Letter for bringing the study to my attention.

Monitoring FDA-Approved Drugs
The researchers examined the safety record of all new drugs approved by the U.S. Food and Drug Administration from 2001 through 2010, following up through February 2017.

The post-product release safety issues they tallied included (1) product withdrawals due to safety concerns; (2) FDA-issued “boxed warnings”--warnings added to the label or to a product’s package insert; and (3) FDA-issued safety communications.
 

About a third of the drugs FDA
approved from 2001 to 2010 had
safety issues after coming to
market. (multiple websites)

They found the FDA had approved 222 new drugs (183 pharmaceuticals and 39 biologics) during the 10 year period. Of these, 32% were affected by post-product release safety issues over a median follow-up period of 11.7 years. There were 123 safety events--3 withdrawals, 61 boxed warnings and 59 safety communications. The average time to the first safety event was about 4 years.

Analyzing the statistics further, the researchers determined the safety events were significantly more frequent among biologics, drugs for treatment of psychiatric disease and drugs that received accelerated and near-regulatory deadline approval.

Wrap Up
For the researchers, the findings highlighted the need for continuous monitoring of new drug safety throughout the drug’s life cycle. The Berkeley Wellness Letter article focused more on the importance of considering all the options before switching to a new drug.

Is the improvement offered by the new drug worth the increased risk? Is it possible to allow more time for the new drug’s benefits and potential problems to be evaluated? When you see those TV commercials, don’t forget the warnings given at the conclusion are only the initially recognized limitations. Of course, unlike me, you have to have the sound on and be listening. Thanks for stopping by.

P.S.
Drug safety study in Journal of the American Medical Association (JAMA): jamanetwork.com/journals/jama/article-abstract/2625319
 
University of California, Berkeley Wellness Letter, August 2017, “Sometimes it’s good to be a neophobe.” pg 3.

Co-Witness Familiarity

Welcome back. A few weeks ago, when I was blogging about using body camera video transcriptions to address the limitations of citizen recollection and direct observation (Traffic Stop Racial Disparities), I came across a somewhat related study of equal interest. Researchers from the U.K.’s University of Huddersfield had examined how a preexisting relationship among eyewitnesses can influence the witnesses’ statements.

Police interviewing witnesses after
London Bridge terror attack,
3 June 2017. (Photo from New York
Daily News and multiple websites)

Eyewitness Importance and Reliability
Eyewitnesses often provide the major lead in an investigation. Still, of the wrongful convictions exonerated by DNA testing, most--some 70%--were due to misidentifications.

The factors behind the current study of misidentifications are well-established:
- Eyewitnesses can be influenced into false recall by co-witnesses to the extent that blame may be wrongfully placed.
- Most eyewitnesses to an event have a preexisting relationship with their co-witnesses.
- People are more inclined to accept the judgement of people they know.

While the effects of co-witness familiarity have been examined for pairs of eyewitnesses to an event, social psychology suggests the effects would be very different when more than two eyewitnesses are involved, which is a common occurrence.

Testing Co-Witness Influence
The current study set out to observe how discussions within groups of eyewitnesses to an event affect blame attribution and, further, to determine if co-witness familiarity has an effect.

Toward these ends, the researchers enlisted 420 participants, average age 33, and separated them into 84 groups of 5 persons. Participants in 36 of the groups had preexisting relationships of at least 3 months; those in 16 groups were strangers to one another; and those in the 32 control groups were also unfamiliar to one another.

The participants all viewed about 1.5 minutes of TV footage of a bar fight. The fight began when a man dressed in green attacked a man dressed in yellow. The two men fought for 40 seconds before being separated.

Participants in all except the control groups then discussed within their groups what they’d seen. Control group participants were not permitted to discuss the event.

Finally, each participant was interviewed individually for his or her statement of what they had witnessed, especially who threw the first punch. They were asked to avoid guessing and report if they were uncertain.

Taking the individual responses of each group’s five members, the researchers calculated the percentage of members in agreement (average statement similarity) and the percentage of members who blamed the correct suspect (blame attribution accuracy). For example, if a group had 3 of 5 members in agreement, the average statement similarity would be 60%, but if those 3 members agreed on the wrong suspect, the blame attribution accuracy would be 2 of 5, or 40%.

Wrap Up
Groups with preexisting relationships had significantly higher levels of both statement similarity and blame attribution accuracy than either the no-relationship or control groups. There was no significant difference in similarity or accuracy between the latter two groups, suggesting post-event discussion among strangers had no real influence.

Of note regarding blame accuracy is that the percentage of eyewitnesses who were uncertain in groups with preexisting relationships was significantly lower than in the other groups, lending further support to eyewitnesses being more susceptible to influence by co-witnesses they know.

Given that co-witnesses with preexisting relationships pose the highest risk of influencing each other’s statements, the researchers emphasize the importance of determining if eyewitnesses have discussed the event prior to giving statements and if the witnesses know one another. Police officers can attempt to assist eyewitnesses to differentiate between witnessed and post-event information.

Try to listen for comments about co-witness discussion and familiarity the next time you’re selected for jury duty. Thanks for stopping by.

P.S.
Study presented at Division of Forensic Psychology British Psychological Society Annual Conference, 13-15 June 2017: Investigating the Effects of Pre-existing Co-Witness Relationships on Statement Similarity. eprints.hud.ac.uk/id/eprint/32263/

Article on study on ScienceDaily website: www.sciencedaily.com/releases/2017/07/170706134649.htm

Don’t Want to Know

Welcome back. Did you happen to see my blog post of a few weeks ago, Political Fact Checking? I summarized a study of how partisanship and prior beliefs influence the way people process political misinformation, specifically Donald Trump’s false statements during the presidential primaries. The results suggested that politicians can spread misinformation without losing supporters.

Avoiding information. (Modified
from photo on multiple websites)

I ask if you saw the post, because there’s a recent study I thought provided further insight on the topic. Investigators from Carnegie Mellon University examined how and why people purposely avoid information as well as some of the consequences. 

Their assessment, which was based on review of research published in economics, psychology and other fields, focused on active avoidance--when the individual is aware the information is available and either has avoided or would avoid free access to it.

Information Avoidance Methods
The investigators categorized the principal tactics used to avoid information, presented here with examples.

   Physical Avoidance -- avoiding certain newspapers, TV or radio shows or conversations with specific people; not returning for results of medical tests (e.g., HIV AIDS).
   Inattention -- seeing a headline and deciding not to read the article, or reading the article and choosing not to think about it.
   Biased Interpretation -- weighing and interpreting information in a manner that supports what they believe and denigrating evidence that contradicts their beliefs.
   Forgetting -- deliberately and selectively failing to review negative information and, in time, forgetting it. (Notably, forgetting may help people deal with bad experiences.)
   Self-Handicapping -- choosing tasks that are too easy or too difficult or taking actions that undermine their performance to avoid information about their own abilities.

Information Avoidance Consequences
I doubt you would have any difficulty coming up with a list of consequences of avoiding information. The effect on decision making topped the investigators’ and probably anyone’s list, whether it’s not reading calories on a label or potentially useful feedback.

Media bias is another significant consequence. Media outlets have incentive to provide biased coverage that aligns with the perspective of their target audience. Nowadays it’s easy to load up on information while avoiding perspectives that challenge one’s existing views.

Among other consequences the investigators discuss are groupthink, when people adopt the shared belief rather than collect their own information; spread of disease could occur if people avoid being tested out of fear they have a contagious disease, or the related ethical transgression if one avoids being tested so as to not confront the dilemma of sharing bad new the test might reveal; and climate change denial, where rejection of a near-unanimous scientific consensus almost by definition requires information avoidance.

Wrap Up
People avoid information that threatens their happiness and wellbeing, which are intertwined with their beliefs. “Not knowing” proffers plausible deniability; knowing might make them feel bad.

Thinking about the earlier blog post on political fact-checking, I would expect those who did not support Trump during the primary would have fact-checked much of what he said because the results often made them happy.

In contrast, I would expect Trump supporters to have avoided fact-checking, certainly, because they trusted their candidate, but also because discovering he was wrong would have made them feel bad. Of course, there had to be those who just didn’t care.

Thanks for stopping by.

P.S.
Carnegie Mellon study in Journal of Economic Literature:
www.aeaweb.org/articles?id=10.1257/jel.20151245
Article on study on ScienceDaily website: www.sciencedaily.com/releases/2017/03/170310121732.htm

Tweet?

Welcome back. Do you tweet or follow anyone who does? I’ve been told that I should open a Twitter account and tweet away. After hemming and hawing and giving profound consideration to the possibility for a year or two, I’ve come to the conclusion that I’m not sure. (You may recall that decisions aren’t my forte, Decision-Making Time.) 
 

Oops. Wrong bird for Twitter
 symbol. (photo from one on
 multiple websites)

If you’ve also delayed joining the Twitter universe or are new to thinking about it, you might be interested in the list of pros and cons I’ve compiled--at least the pros. The cons are more personal. You’re welcome to borrow any.

To Tweet or Not

Pro (P): Everyone is doing it--over half a billion people. Although 40% of Twitter users only read other peoples’ tweets, there are still over 9000 tweets per second.
Con (C): I don’t feel needed.

P: More and more celebrities and politicians are tweeting.
C: (Censored)

P: Twitter is an excellent way to share all the interesting things I do.
C: If I started tweeting, the global economy would be threatened when my expanding network of Twitter followers demanded more and more time off from work to read, retweet (RT) and translate (TT) my daily dose of the interesting things I do.

P: It’s like Facebook but better.
C: It’s like Facebook.

P: Twitter is a way to share my thoughts quickly.
C: I’d have to stop everything I’m doing for a week to count the times I’ve slipped, offended, gotten into trouble or otherwise bungled the message by sharing my thoughts quickly (i.e., with too little thought).

P: I could use Twitter to promote my blog.
C: The Retired--Now What? Blog is unpromotable; too many topics.

P: I would finally be an adopter of more social media technologies.
C: Huh?

P: I could use the same abbreviations I use when I text.
C: I’ll definitely remember that if I ever start texting.

P: Twitter is perfect for mobile phones.
C: And exactly how does that pertain to me?

P: I could learn what people are saying about almost any topic.
C: Whoa! Learning what Twitter users are saying about a topic is probably reliable if I’m keeping up with Lady GaGa and useful if I’m monitoring a disaster or crisis. But Twitter users are only a subset of the population of “people,” and those who tweet aren’t a random sample of that population. Worse, think of the time I’d need to read the tweets about a topic, especially about Lady GaGa, who has over 40 million followers.

P: Since I’m retired, it would give me something to do.
C: Since I’m retired, I’ve got better things to do.

P: News networks are tweeting breaking news.
C: Oh, great! More cryptic headlines and less news.

P: It would force me to write concisely.
C: A limit of 140 characters, including spaces, isn’t really that concise. With my technical writing background, I would have no difficulty con…

Wrap Up

Twitter came to mind because it recently filed for an initial public offering, but there are so many other possibilities. How about Tumblr, where my blogs could be micro? (Lady GaGa uses Tumblr.) Kik is out; I’d need a smartphone. Shapchat is all about photos and it’s too ethereal. I’m way too outdated for Pheed. As always, your kibitzing would be greatly appreciated. Use as many characters as it takes.

Thanks for stopping by.

P.S.

Twitter website and statistics:
twitter.com/
www.statisticbrain.com/twitter-statistics/

Help for potential Twitter users:
techland.time.com/2013/05/21/twitter-101-understanding-the-basics/?xid=newsletter-techland
www.jhische.com/twitter/
thenextweb.com/twitter/2012/09/15/a-list-twitters-language/
mashable.com/2013/07/19/twitter-lingo-guide/
techland.time.com/2013/09/13/5-surprising-things-you-can-find-on-twitter/?xid=newsletter-techland

Wikipedia’s list of social networking sites (not current):
en.wikipedia.org/wiki/List_of_social_networking_websites

Decision-Making Time

Welcome back. How are you at making decisions? I’m big on intuition (ask Myers or her mother, Briggs); but I’m kind of wishy-washy if I don’t have a mountain of information and the results of surveying at least half the people on the East Coast. 
 
Retired, I’m having a terrible time deciding what to do first. Life was so much easier before I retired. Go to work early, come home late, do what I could in the evening and on weekends, feel content. Now, by the end of the day, it’s what I could’a, should’a, would’a done.

Finally, a chance to use one of my sunset photos.

Thinking back, I’m grateful that when I had to make one my biggest decisions, one that changed my life, I had help from above. 
 
Oh, not that above, just the ionosphere.
 

Indecision
 
I don’t remember if I’ve always been this indecisive. In the years before I retired, I was fortunate to team with a colleague who saw my gray areas as black and white. I didn’t necessary adopt his recommended decisions, but at least I had his unambiguous answers as a cushion.
 
I usually did follow his advice on the simple stuff. “Should we cancel the meeting? It’s supposed to snow.” What a relief to hear, “No, absolutely not.” Never mind that when winter weather threatened, he drove a 4-wheel-drive monster that was big enough to push snowplows out of his way.
 
He enjoyed attributing our decision-making differences to my academic and his military officer backgrounds, though I doubt he believed that. Anyway, I’ve known many in academia who, unlike me, never met a decision they had to dawdle over.  

My Big Decision
 
Getting back to the biggest, life-changing decision I ever agonized about, I had to decide whether or not to leave academia.

I had spent a year’s sabbatical leave in the Washington, DC, area, and I loved the change. A couple of years later, after I returned to academic life, an opportunity arose to switch careers and move back to the Washington area.  

Lining up the pros and cons with my wife, who is rarely decision-challenged, didn’t produce any clear answer. One never knows all the pros and cons, but in some cases, we weren’t even sure which was which.

I was ready to flip a coin until I was driving late one night, listening to the radio.

Different car, radio and time, but you get the idea.

In the Washington area, I used to listen to a 24-hour news-weather-traffic report radio station on the AM dial. In my academic habitat, over 300 miles away, I listened to a local station that broadcast at nearly the same frequency. 

Driving that night, I lost the local station. Trying to tune it in again, I was stunned to hear the Washington traffic report. I knew the report was being delivered by a propagating signal, not divine guidance. Still, when the ionosphere speaks, you’ve got to listen. And I never regretted the decision.

Wrap Up

Forced to live with our son’s cat, Henry, since last summer, I’ve struggled to understand his decision-making process and ability. I’ve no particular interest in ethology (i.e., animal behavior); it’s purely defensive on my part. I’ll tell you about it next Friday.

Thanks for stopping by. I think you'll like what's planned for next Tuesday's photo addendum.