[Paper Review] The QWERTY effect on the web: How typing shapes the meaning of words in online human-computer interaction

David García, Markus Strohmaier|arXiv (Cornell University)|Apr 8, 2016

Digital Communication and Language38 references3 citations

TL;DR

This study investigates the QWERTY effect—where right-hand keyboard letters are subconsciously linked to positive meanings—across 15 large-scale web datasets. Using statistical analysis of word usage in titles and reviews, it finds that texts with more right-side letters are rated more positively during both decoding (evaluation) and encoding (writing), with effect sizes of 1–3% of standard deviation in ratings, demonstrating a robust, subconscious influence of keyboard layout on online sentiment perception.

ABSTRACT

The QWERTY effect postulates that the keyboard layout influences word meanings by linking positivity to the use of the right hand and negativity to the use of the left hand. For example, previous research has established that words with more right hand letters are rated more positively than words with more left hand letters by human subjects in small scale experiments. In this paper, we perform large scale investigations of the QWERTY effect on the web. Using data from eleven web platforms related to products, movies, books, and videos, we conduct observational tests whether a hand-meaning relationship can be found in decoding text on the web. Furthermore, we investigate whether encoding text on the web exhibits the QWERTY effect as well, by analyzing the relationship between the text of online reviews and their star ratings in four additional datasets. Overall, we find robust evidence for the QWERTY effect both at the point of text interpretation (decoding) and at the point of text creation (encoding). We also find under which conditions the effect might not hold. Our findings have implications for any algorithmic method aiming to evaluate the meaning of words on the web, including for example semantic or sentiment analysis, and show the existence of "dactilar onomatopoeias" that shape the dynamics of word-meaning associations. To the best of our knowledge, this is the first work to reveal the extent to which the QWERTY effect exists in large scale human-computer interaction on the web.

Motivation & Objective

• To test whether the QWERTY effect—linking right-hand keyboard usage to positive sentiment—manifests in real-world web interactions.
• To examine if the effect occurs not only during text interpretation (decoding) but also during text creation (encoding).
• To identify boundary conditions under which the QWERTY effect may weaken or reverse, such as in high-sales or low-frequency-word contexts.
• To assess the impact of linguistic and contextual variables on the strength of the hand-meaning association in online content.
• To evaluate the implications of the QWERTY effect for sentiment analysis and semantic modeling in computational social science.

Proposed method

• Measured the Right Side Ratio (RSR) as the proportion of right-hand letters in words, using the standard QWERTY keyboard layout.
• Applied permutation tests, robust regression, and multivariate controls to isolate the effect of RSR from linguistic and contextual confounds.
• Analyzed eleven datasets of user evaluations (e.g., IMDB, Amazon, YouTube) to test the decoding effect on item ratings and likes.
• Analyzed four product review datasets with star ratings to test the encoding effect, linking RSR to evaluative sentiment.
• Controlled for word frequency, length, and semantic content to isolate the influence of hand usage on perceived meaning.
• Explored interactions between RSR and contextual factors such as product popularity and word infrequency.

Experimental results

Research questions

• RQ1 Does the QWERTY effect manifest in real-world web interactions, where users evaluate items by rating or liking them?
• RQ2 Is the QWERTY effect detectable not only during interpretation of text (e.g., judging movie titles) but also during the creation of text (e.g., writing reviews)?
• RQ3 Under what conditions does the QWERTY effect weaken or disappear, such as in high-sales or low-frequency-word contexts?
• RQ4 How strong is the effect relative to other psycholinguistic influences, and does it persist after controlling for linguistic and contextual variables?
• RQ5 Can the Right Side Ratio serve as a predictive feature in sentiment analysis models?

Key findings

• The QWERTY effect is robustly observed across 15 web datasets, with a normalized coefficient of 1–3% of the standard deviation in user evaluations per unit increase in Right Side Ratio (RSR).
• For every 100 additional left-hand letters in a review, the star rating decreases by up to 17%, indicating a significant negative association with evaluative sentiment.
• The effect is stronger in low-frequency words and diminishes in high-sales products, suggesting popularity may moderate the effect.
• An inverse QWERTY effect was observed in some contexts, indicating that contextual properties can reverse or suppress the typical hand-meaning association.
• The QWERTY effect is detectable not only in decoding (e.g., movie titles on IMDB) but also in encoding (e.g., Amazon reviews), confirming its presence in both user evaluation and content creation.
• The effect is small in magnitude (1–3% of standard deviation) but consistent with other psycholinguistic effects, suggesting a subconscious, embodied influence of keyboard layout on meaning perception.