Cognitive ability and political preferences in Denmark Emil O. W. Kirkegaard 1 Julius D. Bjerrekær 2 Noah Carl 3 29. August 2016 1. Introduction The relationship between cognitive ability and political opinions is contested in general (Carl, 2014a; Ganzach, 2016; Onraet et al., 2015; Solon, 2014). Several recent studies have examined the relationship between cognitive ability and two axes of political opinion: 1) personal liberty and economic liberty (non-interventionalism/free-marketism). Prior studies have found positive relationships between both axes of freedom and cognitive ability, but negative correlations between the two axes (Carl, 2014b, 2015). These studies have been carried out in the UK and the US, and it is currently unknown whether the findings will hold in other countries such as Denmark. The present study aims to determine whether this is the case. Furthermore, the sample already contains self-reported levels of personal liberalism and economic liberalism and it is of interest to find out how well these correlate with the traits estimated from 10 questions. The study builds upon a previously used sample with N 500 and many variables including a brief measure of cognitive ability (previous study: Kirkegaard & Bjerrekær, In review, cognitive test: 2016). It is possible to gather more data on the same sample by linking the unique IDs provided by the pollster across datasets. It is unknown what the retention rate will be, but it is not expected to be below 60% which would result in a sample size of about 300. This survey adds another 23 content variables to the dataset: 10 about each political axis and 3 additional cognitive items. Due to disagreement about whether using 1-7 or 0-100 Likert scales is best, we decided to gather half the data using each and compare. Emil prefers 0-100 due to less violation of continuity, Noah prefers 1-7 due to more common use/familiarity. A prelimenary explorative study will be undertaken. Planned analyses: 1. Examine how similar the correlation matrices are for the two scales when analyzed using latent 1 Research fellow at Ulster Institute for Social Research. Email: emil@emilkirkegaard.dk 2 University of Aalborg, Denmark. Email: julius.d.b@icloud.com 3 Nuffield College, Oxford. Email: noah.carl@nuffield.ox.ac.uk
and ordinary correlations. 2. Which way it is best to combine the data at the case-level: 1. By converting 1-7 to the midpoints of equal sized bins of 0-100. 2. By convert 1-7 to equally distanced points between 0-100. 3. The method which when used to combine the data at the case-level results in the smallest discrepancy between the combined dataset and the mean correlation matrix of the two datasets analyzed individually (using latent correlations and correlations, respectively). The data can then be used together for the rest of the study. The questions will be given to the participants in Danish, but English versions can be found below as well. 2. Questions English 2.1. Personal liberty 1. A woman should have a legal right to an abortion if she wants one for any reason. 2. The government should punish people who utter racist opinions. 3. It should be legal to utter blasphemous opinions. 4. It should be legal to use, sell and manufacture cannabis and related products. 5. It should be prohibited to use, sell and manufacture LSD and related products. 6. It should be illegal to buy and sell sex. 7. Burning the Danish flag (Dannebrog) should be made a punishable offense. 8. It should be legal for adults to ride motorcycles without using a helmet. 9. It should be legal for adults to ride a car without using a seat belt. 10. The government should restrict citizens access to pornography on the internet. 2.2. Economic liberty 1. It's not the role of the government to make sure that everybody can get a job. 2. The government should mandate a higher minimum wage. 3. The government should not protect domestic industries from competition from other countries. 4. The government should mandate a maximum salary. 5. Taxes on income should be cut.
6. Taxes on property should be increased. 7. The government should keep regulations on business to a minimum. 8. The government should reduce income differences between the rich and the poor. 9. Large companies should be controlled by the state, not private actors. 10. The taxi law should be reformed such that there are fewer rules and companies like Uber become legal. 2.3. Cognitive items Integer input. 1. A bat and a ball cost $1.10 in total. The bat costs $1.00 more than the ball. How much does the ball cost? 2. If it takes 5 machines 5 minutes to make 5 widgets, how long would it take 100 machines to make 100 widgets? 3. In a lake, there is a patch of lily pads. Every day, the patch doubles in size. If it takes 48 days for the patch to cover the entire lake, how long would it take for the patch to cover half of the lake? Multiple choice. 4. Suppose you live in a place that has a constant chance of being struck by lightning at any time throughout the year. Suppose that the strikes are random: every day the chance of a strike is the same, and the rate works out to one strike a month. Your house is hit by lightning today, Monday. What is the most likely day for the next bolt to strike your house? Tuesday, the day after. Wednesday, 2 days later. Friday, 4 days later. Monday, next week. Money, next month. Tuesday, next month. Every day has the same chance. [order of answers randomized] Copied from Pinker (2012).
3. Questions Danish 3.1. Personlig frihed Translations are ours. 1. Kvinder bør have ret til abort uanset årsag. 2. Regeringen bør straffe personer som ytrer racistiske holdninger. 3. Det bør være lovligt at ytre blasfemiske holdninger. 4. Det bør være lovligt at anvende, sælge og fremstille cannabis/hash og relaterede produkter. 5. Det bør være forbudt at anvende, sælge og fremstille LSD og relaterede produkter. 6. Det bør være forbudt at købe og sælge sex. 7. Det bør være strafbart at brænde det danske flag (Dannebrog). 8. Det bør være lovligt for voksne at køre motorcykel uden styrthjelm. 9. Det bør være lovligt for voksne at køre bil uden at bruge sikkerhedssele. 10. Regeringen bør begrænse borgernes adgang til porno på internettet. 3.2. Økonomisk frihed 1. Det er ikke regeringens opgave at sørge for at alle kan få et job. 2. Regeringen bør indføre en højere minimumsløn. 3. Regeringen bør ikke beskytte danske industrier mod udenlandsk konkurrence. 4. Regeringen bør sætte en grænse for hvor meget man kan få i løn. 5. Indkomstskatten bør reduceres. 6. Ejendomsskatten bør forøges. 7. Regeringen bør holde reguleringer af virksomheder til et minimum. 8. Regeringen bør reducere indkomstforskellene mellem rige og fattige. 9. Store virksomheder bør kontrolleres af staten, ikke private aktører. 10. Taxaloven bør reformeres således at der bliver færre regler og virksomheder som Uber bliver lovlige. 3.3. Kognitive items Integer input. 1. Et bat og en bold koster i alt 110 kr. Battet koster 100 kr. mere end bolden. Hvor meget koster
bolden? 2. Hvis det tager 5 maskiner 5 minutter at lave 5 dingenoter, hvor mange minutter ville det så tage 100 maskiner at lave 100 dingenoter? 3. Der er en gruppe åkander i en sø. Hver dag bliver gruppen dobbelt så stor. Hvis det tager 48 dage for gruppen at dække hele søen, hvor mange dage ville det så tage for gruppen at dække halvdelen af søen? Multiple choice. 4. Forestil dig at du lever i et hus som har en konstant chance gennem året for at blive ramt af et lyn. Antag at lynnedslag er helt tilfældige: chancen er den samme hver dag og lynet slår ned i dit hus i gennemsnit en gang om måneden. Lynet slog ned i dit hus i dag, mandag. Hvilken dag har størst sandsynlighed for at blive den næste dag dit hus bliver ramt af et lyn? Tirsdag, dagen efter. Onsdag, 2 dage efter. Fredag, 4 dage efter. Mandag, ugen efter. Mandag, måneden efter. Tirsdag, måneden efter. Alle dage har den samme sandsynlighed. [Tilfældig rækkefølge] 4. Predictions and planned analyses Based on previous findings and our own judgment calls, we predict that: 1. There will be a common factor for each of the two scales with none or few (<3) items not loading on the common factor in the correct direction. 2. Cognitive ability will be positively related to the freedom end of both axes. The effect sizes will be r =.25 ±.15. 3. The two political axes will be negatively related. Effect size will be r = -.35 ±.15. 4. Self-rated and measured personal liberalism will correlate poorly. Effect size will be r =.20 ±. 30. 5. Self-rated and measured economic liberalism will correlate moderately. Effect size will be r =. 40 ±.30. 6. There will be a non-linear relationship between cognitive ability and the economic liberalism scale such that the positive slope will decline in strength or perhaps reverse near the end of the ability axis.
All effect size predictions concern values corrected for measurement unreliability in both variables if possible. 4.1. Planned analyses We plan on doing the below analyses. These will be conducted in R. Corrections for measurement error are done using Spearman's formula. 4.1.1. Cognitive ability Cognitive ability is measured as the latent factor of the 8 total cognitive items: the ICAR5 and the 3 CRT items. This is done via IRT factor analysis. 4 Compare scores from IRT and unit weighted scoring. 4.1.2. Political scales Use factor analysis on the two political scales. 5 Extract scores via factor analytic scoring and unit weights (with reversing of negative items) and compare them. 4.1.3. Predicting political axes Correlate and plot the relationships between cognitive ability and the political axes scores. Fit lines with OLS and LOESS. Regress the two political axes on cognitive ability, age, gender and educational attainment using OLS regression and LASSO regression. 4.1.4. Political axes and party-agreement Correlate agreement with each of the major parties and the two political axes. Calculate mean political axes scores for supporters ( 80% agreement) of the major parties and compare with results from the Stereotype study. Plot the parties on a scatterplot. References Carl, N. (2014a). Cognitive ability and party identity in the United States. Intelligence, 47, 3 9. http://doi.org/10.1016/j.intell.2014.08.003 Carl, N. (2014b). Verbal intelligence is correlated with socially and economically liberal beliefs. Intelligence, 44, 142 148. http://doi.org/10.1016/j.intell.2014.03.005 Carl, N. (2015). Cognitive ability and political beliefs in the United States. Personality and Individual Differences, 83, 245 248. http://doi.org/10.1016/j.paid.2015.04.029 Ganzach, Y. (2016). Cognitive ability and party identity: No important differences between Democrats and Republicans. Intelligence, 58, 18 21. http://doi.org/10.1016/j.intell.2016.05.009 4 We will use irt.fa from Psych package. This is based on the 2PN model (Revelle, 2016). 5 We will use fa from Psych package.
Kirkegaard, E. O. W., & Bjerrekær, J. D. (In review). Are stereotypes about immigrants accurate in Denmark?: a large, preregistered study. Open Differential Psychology. Retrieved from http://openpsych.net/forum/showthread.php?tid=287&pid=4122 Kirkegaard, E. O. W., & Bjerrekær, J. D. (2016). ICAR5: a 5-item public domain cognitive test. Open Differential Psychology. Onraet, E., Van Hiel, A., Dhont, K., Hodson, G., Schittekatte, M., & De Pauw, S. (2015). The Association of Cognitive Ability with Right-wing Ideological Attitudes and Prejudice: A Metaanalytic Review. European Journal of Personality, 29(6), 599 621. http://doi.org/10.1002/per.2027 Pinker, S. (2012). The better angels of our nature: why violence has declined. New York: Penguin Books USA. Revelle, W. (2016). An introduction to psychometric theory with applications in R. Retrieved from http://www.personality-project.org/r/book/ Solon, I. S. (2014). How intelligence mediates liberalism and prosociality. Intelligence, 47, 44 53. http://doi.org/10.1016/j.intell.2014.08.009