Presentation is loading. Please wait.

Presentation is loading. Please wait.

1 Using R for consumer psychological research Research Analytics | Strategy & Insight September 2014.

Similar presentations


Presentation on theme: "1 Using R for consumer psychological research Research Analytics | Strategy & Insight September 2014."— Presentation transcript:

1 1 Using R for consumer psychological research Research Analytics | Strategy & Insight September 2014

2 2 Agenda What this presentation will cover About me 1 Analytics Challenge 2 Standard Analytics 3 Quantitative Psychology 4 Artificial General Intelligence 5 Conclusion 6

3 3 About me

4 4 About me Background & experience

5 5 Analytics Challenge

6 6 Analytics Challenge Transition KM from SPSS to R

7 7 Analytics Challenge Change management & user aids rcommander.com

8 8 Standard Analytics

9 9 Standard Analytics An overview of a subset of common techniques Tools

10 10 Quantitative Psychology

11 11 Quantitative Psychology Quantitative consumer psychology in a nutshell Quantitative consumer psychology specializes in measurement, methodology and research design and analyses relevant data to better understand and (hopefully) predict consumer cognition and behaviour psych: Procedures for Psychological, Psychometric, and Personality Research IAT: Functions to use with data from the Implicit Association Test Useful packages in R: StatMatch: Integrating two or more data sources Initial developments in R Computational Cognitive Modelling

12 12 Traditional Psychometrics [Package ‘psych’]

13 13 Package ‘psych’ Numerous standard psychometric analyses at your fingertips Author and maintainer: Prof. William Revelle | revelle@northwestern.edu Scale construction using factor analysis, cluster analysis, and reliability analysis Graphical models (e.g., SEM) for path diagrams to explore and test theories Perceived Usefulness External Variables Perceived Ease of Use Attitude Toward Using Behavioural Intention to Use Actual Use

14 14 Implicit Association Test [Package ‘IAT’]

15 15 Overview of IAT What is this good for? The IAT is a computerised method for indirectly measuring the strength of the association via a double-categorisation task Psychological assumption: People are able to categorise strongly associated concepts more quickly than concepts that are weakly associated (Greenwald et al., 1998) The IAT effect is usually interpreted as a measure of implicit attitude

16 16 IAT: Experimental details How does it work? Science or Male Liberal Arts or Female Engineering Science or Male Liberal Arts or Female Mike Liberal Arts or Male Science or Female Engineering Liberal Arts or Male Science or Female Tiffany congruent incongruentThe difference (D-Scoring algorithm) in response latencies between the two concepts for the congruent and incongruent pairing provides the basis for the IAT measure (Greenwald, 2003) Congruent pairing Incongruent pairing

17 17 Package ‘IAT’ Measuring hidden cognitive associations Author and maintainer: Dan Martin | dpmartin42@gmail.com Implements the standard D-Scoring algorithm (Greenwald et al., 2003)

18 18 Data Integration [Package ‘StatMatch’]

19 19 Package ‘StatMatch’ Data Integration / Data Fusion The availability of data is often a serious problem Data fusion provides a way out by combining information from different sources into a single data set The algorithm typically consists of three steps. I.Decide on the direction of the fusion and the purpose of the fusion II.The matching distance is calculated over some subset of the common variables III.Loop through datasets and assign donors based on distance match and penalty weight

20 20 Package ‘StatMatch’ Data Integration / Data Fusion Statistical matching integrates different data sources in order to investigate the relationship among variables not jointly observed in a single data source

21 21 Computational Cognitive Modelling

22 22 Computational Cognitive Modelling What is this?

23 23 Computational Cognitive Modelling Schematic overview

24 24 Soar: Artificial General Intelligence (AGI)

25 25 AGI using R? rrules Package ‘rrules’ General rule engine: Lisp- like rule processing MDPtoolbox Package 'MDPtoolbox' Reinforcement learning (QL) neuralnet Package 'neuralnet' Neural networks

26 26 Conclusion The future for R in Quantitative Psychology looks promising R has the potential to replace SPSS due to it’s flexibility, power, and increasing availability of packages There are still several significant challenges to R becoming mainstream (at least in psychology) Computational modelling is critical for mechanistically understanding cognition The R language is still largely underexplored in computational cognitive modelling

27 27 Thank you!


Download ppt "1 Using R for consumer psychological research Research Analytics | Strategy & Insight September 2014."

Similar presentations


Ads by Google