Automating Implicit Motive Coding

This web application has been developed by Hiram Ring (PhD, Linguistics) and Joyce S. Pang (PhD, Psychology) to classify English text according to the presence of three implicit human motives, described as (the needs for) power, achievement, and affiliation (nPow, nAch, nAff).1

The particular neural network model used here (Convolutional Neural Network; CNN)2 achieves moderate but highly significant correlations with human-coded scores on a range of datasets, as well as showing various forms of validity with such datasets (see Pang & Ring, 2020).3 As such, while the underlying model represents the current state of the art in automated motive coding (for English), it cannot replace a human coder.

This application is intended to demonstrate the potential use for automated motive coding, but should not be used for assessment or diagnostic purposes. Research toward the automation of implicit motive coding is ongoing and this website will be updated with new models that achieve higher correlations with human coders.

The application is limited to predicting motive codes for five sentences only. Full paragraphs of text can be pasted into the text box, which will then be split into sentences, of which only the first five will be coded. If you are interested in comparing the model's predictions with your own human-coded scores on a larger dataset, would like to provide data that could improve future models, or are otherwise interested in this work, please contact the authors at > iminfo AT implicitmotives D@T com <.

  1. The handbook Implicit Motives contains helpful information about the research concerning these human motives, including the process for assessing them using human coders.

  2. For more information on CNN and other Deep Learning architectures, this is a good resource.

  3. Read the published article here. For the author accepted manuscript, data, and other supplementary material, visit the osf page.


Pang, J. S., & Ring, H. (2020). Automated Coding of Implicit Motives: A Machine-Learning Approach. Motivation and Emotion, 44(4), 549-566. DOI: 10.1007/s11031-020-09832-8. Both authors contributed equally to this work.

Copyright © 2020