Internet and computer research
Presentations Smart phones Virtual reality
Papers Due tonight! Any questions?
Examples of internet research Why is it used? Why should it be? When should it NOT be? Examples of Translational Phenomenological Novel
Ways to recruit online (and issues?) Standard populations Respondent-driven sampling Targeted populations Crowdsourcing (we’ll come back to this one)
Issues with online studies What are some issues/potential problems? Any solutions for these? What are advantages to doing studies online? What about advantages and disadvantages to big data? http://www.outofservice.com/
What is the future of internet research? Big data—what are some particular issues with this method? Smartphones
mTurk Issues with samples Motivation Demographics Non-naïve Social desirability Cross-talk
Arguments for/against IMCs
Necka, Cacioppo, Norman, & Cacioppo, 2016 Compared mTurk to campus (not just intro) and community (panel) participants Reported on how often they or others did behaviors Issues/thoughts on the study?
Differences between groups (Necka et al., 2016) Multitask Look things up Lie on age and gender Look for researchers they know Table 2—other interesting findings? How much are these results due to lab vs. online?
To get better data Illustrate that important study Make interesting
Facebook data How are Facebook and other OSNs used in research? Should we be worried or excited or both? What does Facebook research tell us that’s new?
Fig. 1. Facebook users and articles: Cumulative totals by year. Wilson, Gosling, & Graham. Perspectives on Psychological Science 2012;7:203-220 Copyright © by Association for Psychological Science
Table 1. Areas of Facebook Research Identified in the Literature Review. Wilson, Gosling, & Graham. Perspectives on Psychological Science 2012;7:203-220 Copyright © by Association for Psychological Science
Table 2. Number of Articles Published by Category and Year . Wilson R E et al. Perspectives on Psychological Science 2012;7:203-220 Copyright © by Association for Psychological Science
T3 study (Lewis et al., 2008) Social network analysis What types of data did they collect/code? What about the measures of ethnicity and SES? Who/what are Facebook friends? Which type of internet study would this be? https://dataverse.harvard.edu/dataverse/t3
More T3 Is it okay to get this type of data? To make it available? What are the main types of problems with this type of study? Positives? How could spatial stats be used in a study like this? What are some results from this study? (tables, p. 337)
Further research http://www.pnas.org/content/109/1/68.full
Parameter estimates β and 95% confidence intervals for selection and influence effects from 15 models of the coevolution of friendships and tastes (n = 211 for music, n = 201 for movies, n = 191 for books). Parameter estimates β and 95% confidence intervals for selection and influence effects from 15 models of the coevolution of friendships and tastes (n = 211 for music, n = 201 for movies, n = 191 for books). Significant coefficients are labeled with an asterisk, where a coefficient is considered significant if the 95% confidence interval does not contain β = 0. Selection effects measure the tendency for a tie to develop between two students who both express tastes in the given cluster; influence effects measure the tendency for students whose friends express tastes in the given cluster to themselves adopt tastes in that cluster (SI Materials and Methods, Coevolution of Tastes and Friendships). Lewis K et al. PNAS 2012;109:68-72 ©2012 by National Academy of Sciences
Next week Class on TUESDAY (12:30) One reading will be emailed to you Others available online 3 presentations next week (come see me)