Word Frequency Effect


Have you ever noticed that you can read a passage in a novel or magazine much more quickly than you can read a passage in a textbook? The word frequency effect, a psychological phenomenon where words we see often (such as 'cat') are recognized more quickly than words we see less often such as 'vat', is probably partly to blame. Novels tend to contain high frequency words people use in everyday speech whereas many textbooks contain new, complex, low frequency words that you may not have seen before, or have seen very rarely. If you have some books handy see for yourself by flipping to a random page; chances are your novel will contain words that you can easily skim over whereas the text will be full of new words that you have to think about a bit before you can understand what they mean. The reason for the word frequency effect is hotly debated, some researchers think that seeing a word more often lower the threshold of activation for that word meaning you need less additional activation to recognize the word when it's presented, other argue that higher frequency words have a higher 'resting level of activation' so that less additional activation is needed for recognition when the word is presented. Whichever model is correct it is certain that any model of word recognition must account for this very robust effect.

Higher frequency word has lower threshold than lower frequency words

Higher frequency words have higher resting level of activation than lower frequency words

Let's see it in action! We will be using the lexical decision task to look at the word frequency effect.


The size of the word frequency effect will depend on stimuli used with the greatest effect observed when the high frequency words are extremely common (such as 'THE') and the low frequency words are almost unheard of (such as 'STRAIT'). The word frequency effect has led cognitive psychologists developing models of word recognition to hypothesize that seeing a word has an impact on some aspect of the way the word is represented. Some researchers argue that seeing a word many times lowers the activation threshold for the word, whereas other researchers argue that high frequency words have a higher resting level of activation (in both case less activation is needed to reach threshold than a low frequency word making recognition quickly). It is possible that high frequency words are connected to so many other words that spreading activation is responsible for the higher recognition speed.

Review Questions

Question: Would you expect the number of errors made in the high and low frequency word condition to be the same in a typical lexical decision? Explain your answer.

Answer: No error rates won't be the same, there will be more errors in the low frequency condition because in some cases the participant will classify an unfamiliar low frequency word as a nonword. In all lexical decision experiments it is very important to only look at response times for correct trials and to look at error rates too. If your error rates are very high in the low frequency word condition then what you are observing is more of a lexicality effect (and effect of the stimulus being a known word or not) instead of an actual word frequency effect.

Question: Suppose you gave one group of participants the list of low frequency words that you were going to use in the experiment to 'study' a week ahead of time, also assume that they actually studied them. Would you expect the word frequency effect to be different in these groups when they were tested? Explain your answer.

Answer: Yes, you would expect a difference because the 'study group' would be highly familiar with the low frequency words from having seen them over and over while studying. The 'study group' would have a much smaller word frequency effect than the 'control' group.