Spot-the-word task for child-directed speech

What's the spot-the-word task?

In this task, the participant (be it a human or a machine) receives two spoken (or written) stimuli that form a minimal pair of (word, non-word). For instance: brick and blick. The participant is then asked to decide which of the two stimuli is the word. If the latter fails, it obtains a score of 0, if it succeeds a score of 1. The accuracy is computed as the proportion of trials for which the participant succeeded in finding the right word.

When considering a machine participant, one has to extract the probability of the stimulus, which is expected to be higher for the word than for the non-word (how to extract this measure of probability is one of the numerous design choice the programmer is faced with)

Examples of stimuli

Word	Pseudo-word	Word	Pseudo-word
hello	lello pello sero dello sello	cookie	kootie koonie roodie rootie boonie

Table 1: Minimal pairs of real and pseudo-words used in the spot-the-word lexical task.

Getting started

1. Install
2. Create the evaluation set
3. Compute the accuracy