Study and sheet notes

De eerste vragen gaan over het artikel:

Brunell, A.B., Fisher, T.D. (2014). Using the bogus pipeline to investigate grandiose narcissism.Journal of Experimental Social Psychology, 55, 37-42. https://doi.org/10.1016/j.jesp.2014.05.015

1. What is the research question of Brunell and Fisher (2014)?
To what extent are grandiose narcissism and other narcissism-related constructs sensitive to bias in reporting?

2. In the article, an experimental manipulation is used.
a. Please explain, in your own words,what a bogus pipeline is.

The bogus pipeline is a technique used by social psychologists to reduce socially desirable answers. Participants are linked to a device that the researchers tell them is a lie detector. As a result, participants have the feeling that they are being checked to see whether they are responding truthfully. It is assumed that this will actually motivate participants to answer truthfully.

b. Describe, in your own words, what the differences are between the experimental conditions in this study.

The idea behind this is that by using the bogus pipeline technique, amongst other things, the measurement of social constructs becomes more valid. Giving socially desirable answers is a systematic error that affects validity.
Bogus pipeline condition: Participants were told that they would be linked to a lie detector during the experiment. Participants completed the questionnaires while they were linked to the bogus pipeline.
Anonymous condition: Participants in this condition were not linked to a lie detector. Participants filled in the questionnaires in a private setting. 
Exposure threat condition: Participants in this condition were not linked to a lie detector. Participants completed the questionnaires, with a survey assistant in the room. This research assistant informed the participants that he/she would pick up the questionnaire afterwards. In all conditions, the participants submitted the questionnaires in a closed box.

De volgende vragen van de voorbereidende opdracht gaan over replicatie, assumpties en type I fouten. 

Replication Explanation  

Assumptions ANOVA

• Homogeneity of variances
• Independence of observations
• Absence of outliers
• Normal distribution
• Measurement level

Assumptions ANCOVA

• Measurement level of the covariate should be continuous
• Independence of the covariate and the treatment effect (no statistical requirement)
• Homogeneity of regression coefficients

Type I error

Alpha = chance that you falsely reject your null hypothesis (type I error).  
Beta = chance that you falsely reject your alternative hypothesis. 
Power = chance that you correct reject the null hypothesis. 

If you do one analysis the probability of making a type I error = alpha. 
Multiple analysis: probability of making a type I error > alpha. 
Experiment wise error rate: a_ew = 1 – (1 – a_pc)^c
         c = the amount of comparisons