Friday, February 26, 2016

A List of Causal Factors and Other Ideas to Consider When Weighing Causal Claims

Let's say we plan to do a social science study or evaluate one or weigh an argument positing causes.

We should consider plausible alternative causal factors left out by a thinker.

I keep a book on my nightstand. When I read or think of a causal factor or other idea I should keep in mind, I sometimes write it in the first few pages of the book.

The list:
IQ, age, war, race, time, luck, laws, wealth, stress, fraud, zest, genes, diet, height, weight, BMI, wealth, income, culture, status, habits, power, gender, famine, crowding, apathy, egoism, altruism, anomie, placebo, nocebo (feeling harm from a harmless treatment), boredom, honesty, ethnicity, weather, fatigue, religion, diversity, seniority, incentives, geography, popularity, patience, hedonism, confusion, standards, novelty, naivete, sadism, nihilism, obeisance, anxiety, isolation, commitment, experience, solipsism, nationality, expectations, misanthropy, proximity, fanaticism, busyness, motivation, scientism, triumphalism, salience, replicability, ruminations, militarism, aggressiveness, suggestibility, masochism, desensitization, optimism, labor unions, obsessions, compulsions, groupthink, misandry, misogyny, careerism, xenocentrism, ethnocentrism, infrastructure, desensitization, self-loathing, nonbelief, corruption, ruthlessness, credentialism, cronyism, nepotism, competition, cooperation, victory, defeat, tax rates, tax targeting, participant compliance rates, street smarts, mortality rates, faulty study design, family size, tokenism, trustworthiness, skill level, arrival dates, leadership quality, overconfidence, underconfidence, one-sidedness, divorce rates, imitating others, democracy, non-democracy, self-pity, self-contradictions, noise levels, exercise levels, body image, employment status, confirmation bias, perceived weakness, career specialty, opportunity costs, pathogen load, educational level, youth bulges, genetic load, abortion rates, free riding, new discoveries, self-control, shared struggles, shared danger, unethical guilt, guilt proneness, shame proneness, false positives, false negatives, mutual causation, shell shock, Stockholm syndrome, witch hunting, misplaced loyalty, alliance quality, organization level, Hamilton's rule, educational field, absence of study controls, age at which a participant practiced a skill, perceived victimization, availability of distractions, feeling needed, perceived equality, unrepresentative sampling (including survivor bias and self-selection effects), ethical character, monocausal fallacies, body language, faulty study controls, voice tone, group polarization, development rates, job mobility, perceived fairness, drug use (including alcohol and smoking), improved police tactics, residential mobility, moral hazard, laziness, task persistence, hormone levels (testosterone, estrogen), crowd psychology, savings rates, monitoring effects (white man effect), study protocol violations, brain size, brain tumors, brain trauma, availability errors, halo effects, wedge issues, purposefulness, happiness, depression, despair, group cohesion, asceticism.

Plus some more: jealousy, anchoring (believing first claim heard on a subject), antinomianism, anthropomorphism, anthropocentrism, moral aestheticism, mechanical failures, math errors, hindsight error (easier to see errors after the event), chronic pain, acute pain, commitment level, conspicuous compassion, expected value, nutrient deficiencies, Duchenne smiles, political stability, single influential individual, repetition effects (believing claims repeated most often), other popularity effects, feeling entitled, celebrity worship, world weariness, legalized bribery, philosophical farsightedness, probability of success, belief in sacredness, small sample sizes, transitory situational factors, hyperbolic discounting, declining marginal utility, other cognitive impairments, enforcement or nonenforcement of standards, level of cognitive dissonance, perceived legitimacy, belief in karma, tit for tat, belief in ontological guilt (blood libel), inaccurate self-reports, tragedy of the commons, improved 911 service, substance poisoning (lead poisoning), publication bias, self-esteem, feeling superior, religious subtypes, the "big five" (openness to experience, agreeableness, neuroticism, extroversion, conscientiousness), framing effects (slanted questions, leading questions, and bad definitions), paradoxical effects, goal related effects, social skills, public policies, teacher quality, historical factors, teacher training, fatalistic beliefs, scaling effects (inaccurate extrapolation), relativistic beliefs, birth control rates, blood sugar levels, the dark triad (narcissism, Machiavellianism, psychopathy), belief in scientism, peer group types, nation-state formation, divide-and-rule, time spent on activities, threshold effects (straw that breaks the back), social pressure, statistical insignificance, media influence, environmental aesthetics, study confederates influencing results, beliefs about hell, high, unstable self-esteem, chips on shoulders, faith in experts, length of employment, beliefs about grace, belief in authoritarianism, treasonous leadership, bait-and-switch, population increases, population decreases, faulty tests to measure alleged factors, study participants dropping out, fear of losing face, various other isms, chips on shoulders, non-parental environments, failure to ignore sunk costs, regression to the mean, ability to delay gratification, no causal relationship exists, perception of decay, follow the money, follow the power, errors due to random chance, belief in plausible deniability, homogeneous unrepresentative sampling, strength of will, time spent using electronics, browbeating and struggle sessions, viewing others as objects, reversing cause and effect, winner-take-all markets, results are statistically significant but too small to be important, swiftness and probability of punishment, monopoly and oligopoly effects, variance in genotypes and phenotypes, ratio of performance IQ to verbal IQ.

Wait.

I'm not done: number of working adults in a family or household, increased survival of big brained c-section babies, increased survival of small hip genotypes, belief in reciprocation, improved test taking skills, belief in exploiting weakness, belief that the feel good end justifies any means, belief that a feel good means justifies the result, ingroup and outgroup beliefs, improved emergency room medicine, economic opportunities in early adulthood, other technological factors, the probability of being caught, plus numerous other genetic and environmental factors I haven't thought of.

Good luck.

Figuring out causes is a pain in the rear.

"Correlation is not cause" is a banal statement. Correlation by itself does not prove cause is a better statement.

Most plausible study factors aren't looked at in most studies. Many studies are designed to fit the scientist's favored conclusions.

No comments: