2010Quarterly Journal of Experimental PsychologyRequires access

Do the effects of subjective frequency and age of acquisition survive better word frequency norms?

Marc Brysbaert, Michael J. Cortese

Open publisher page 131 citations

Abstract

Megastudies with processing efficiency measures for thousands of words allow researchers to assess the quality of the word features they are using. In this article, we analyse reading aloud and lexical decision reaction times and accuracy rates for 2,336 words to assess the influence of subjective frequency and age of acquisition on performance. Specifically, we compare newly presented word frequency measures with the existing frequency norms of Kucera and Francis (1967), HAL (Burgess & Livesay, 1998), Brysbaert and New (2009), and Zeno, Ivens, Millard, and Duvvuri (1995). We show that the use of the Kucera and Francis word frequency measure accounts for much less variance than the other word frequencies, which leaves more variance to be "explained" by familiarity ratings and age-of-acquisition ratings. We argue that subjective frequency ratings are no longer needed if researchers have good objective word frequency counts. The effect of age of acquisition remains significant and has an effect size that is of practical relevance, although it is substantially smaller than that of the first phoneme in naming and the objective word frequency in lexical decision. Thus, our results suggest that models of word processing need to utilize these recently developed frequency estimates during training or setting baseline activation levels in the lexicon.

About this research paper

What this paper is about

Megastudies with processing efficiency measures for thousands of words allow researchers to assess the quality of the word features they are using. In this article, we analyse reading aloud and lexical decision reaction times and accuracy rates for 2,336 words to assess the influence of subjective frequency and age of acquisition on performance. Specifically, we compare newly presented word frequency measures with the existing frequency norms of Kucera and Francis (1967), HAL (Burgess & Livesay, 1998), Brysbaert and New (2009), and Zeno, Ivens, Millard, and Duvvuri (1995). We show that the use of the Kucera and Francis word frequency measure accounts for much less variance than the other word frequencies, which leaves more variance to be "explained" by familiarity ratings and age-of-acquisition ratings. We argue that subjective frequency ratings are no longer needed if researchers have good objective word frequency counts. The effect of age of acquisition remains significant and has an effect size that is of practical relevance, although it is substantially smaller than that of the first phoneme in naming and the objective word frequency in lexical decision. Thus, our results suggest that models of word processing need to utilize these recently developed frequency estimates during training or setting baseline activation levels in the lexicon.

Why it matters

OpenAlex reports 131 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Megastudies with processing efficiency measures for thousands of words allow researchers to assess the quality of the word features they are using. In this article, we analyse reading aloud and lexical decision reaction times and accuracy rates for 2,336 words to assess the influence of subjective frequency and age of acquisition on performance. Specifically, we compare newly presented word frequency measures with the existing frequency norms of Kucera and Francis (1967), HAL (Burgess & Livesay, 1998), Brysbaert and New (2009), and Zeno, Ivens, Millard, and Duvvuri (1995). We show that the use of the Kucera and Francis word frequency measure accounts for much less variance than the other word frequencies, which leaves more variance to be "explained" by familiarity ratings and age-of-acquisition ratings. We argue that subjective frequency ratings are no longer needed if researchers have good objective word frequency counts. The effect of age of acquisition remains significant and has an effect size that is of practical relevance, although it is substantially smaller than that of the first phoneme in naming and the objective word frequency in lexical decision. Thus, our results suggest that models of word processing need to utilize these recently developed frequency estimates during training or setting baseline activation levels in the lexicon.

Key concepts: Word lists by frequency, Age of Acquisition, Lexical decision task, Lexicon, Word (group theory), Psychology, Reading (process), Cognitive psychology

Related papers

Back to paper searchBrowse research topicsOriginal source
Do the effects of subjective frequency and age of acquisition survive better word frequency norms? — Research Paper | ScholarLens