Year One

2021-2022

October

Defined two overarching research questions, which will motivate my research over the next 3+ years:

  1. Can playing AVGs improve cognition? Mixed findings so far, most positive findings come from one lab (Bavelier & Green) with small samples. Does expertise in AVGs differentially impact cognition? My MSc research provided correlational evidence regarding this, but further analysis of this data set and experimental research is necessary.

  2.  How does playing AVGs improve cognition? 'Learning to Learn' could be an explanation, but there is limited theoretical and experimental research on this topic. 

November

Used stricter filtering and re-analysed the MSc data set, which excluded participants from all levels of analysis who did not have full data for the three expertise measures - this resulted in a reduced sample of n = 273. Expertise clustering in the previous analysis was based on seven measures, this was reduced to three: total hours playtime, weekly hours playtime and self-rated expertise. Cluster group analysis suggested four groups would be the best fit, and allocation was as follows: novice = 105, casual = 107, intermediate = 28 and advanced = 33. These changes provided the following results.

Speed in single rule and repetition rule trials were fastest in the Semi/Professional group. However, the casual group were fastest in switch rule trials, and also showed the lowest switching costs and mixing costs. Significant main effects of expertise group were observed for single-rule trials only. 

These results suggest that processing speed, task switching and mixing abilities vary with expertise however, this relationship is non-linear. Experienced players, were fastest and most efficient in terms of their processing speed. However, contrary to previous findings, players with the second least CS: GO expertise (Casual group) performed most efficiently in mixing and switching conditions. These discrepancies may be explained by the larger sample, and range of expertise in the present sample compared to previous studies. Further examination of the literature is necessary to gain a proper understanding of our results.

April

Diffusion modelling is the next step of the analysis, which maps the cognitive processes involved in fast decision-making tasks, and quantifies performance differences across participants. The EZ-diffusion model determines the most psychologically relevant parameters which are: quality of information or drift rate (ν), response conservativeness or boundary separation (α) and non-decision time (Τer). These parameters are estimated based on the inputs of three observed values from our data: mean response times (MRT), variance of response times (VRT) and proportion of correct decisions (Pc).

Data has been pre-processed and there are values for MRT, VRT and Pc for each participant however, further reading is necessary to develop my understanding of the EZ-diffusion model, before I can successfully complete the analysis.

July

Began the planning and data collection  for Study 2

September

Began data analysis of Study 2 and preparation for Confirmation Review

 
cluster_plots.png
 

Personal Development

2021 - 2022

Biggest Success

Improving my coding skills and presenting at the LaP conference in Finalnd. 

Biggest Challenge

Understanding and executing Drift-Diffusion Model (DDM) analysis.

Personally

Achieved a Distinction on my MSc, got engaged to my partner and moved into our first home!

 

Conferences

Psychonomics

November 2021

Online
Attended

Women in Cognitive Science (WiCSW)

November 2021

Online
Attended

Women in Games Careers and Networking Expo

November 2021

Online
Attended

Learning and Plasticity (LaP)

April 2022

Äkäslompolo, Finland
Presented at

Psychonomics

November 2022

Boston, MA, USA
Will present at