Theory, Model, Method
Linguistic data can be collected through experiments in laboratory settings. Experiments are used to collect (statistically relevant) evidence on a specific aspect of performance, which can then be used to test a hypothesis. An advantage of experiments (over naturalistic data) is that various causal factors (independent variables), which might be responsible for a certain effect (dependent variable), can be controlled and thus analysed seperately. A wide range of experimental methods are used, especially in the field of psycholinguistics. Three examples are given below.
(1) Preferential looking task
This method is designed to test young children's comprehension of language. Two alternative pictures appear on two computer screens. Then the children hear a word / an utterance that corresponds to one of the pictures. The experimenter pays attention to (measures) the direction of children's eye gaze on hearing the word / utterance. When children look at the picture that matches the linguistic information, this is taken as evidence in favour of comprehension. (The picture on the right is taken from Hirsh-Pasek and Golinkoff (1991:303)), who used the preferential looking task to test whether children at the one-word stage know that sentences are formed by combining the meaning of the words contained in them.)
(2) Measuring reaction times: Lexical decision task
An experimenter shows subjects groups of four letters on a computer screen. They have to press a button every time the group of letters represents an actual word.
Example: PLIT - BRON - GIVE - ADGE - LIMP - MISH - DASE - READ - GALF - BARK - CLON - STIR - KNOW - GACK
The experimenter measures (in milliseconds)how much time elapses between the letters appearing on the screen and the subject pressing the button when appropriate. Reaction Time to GIVE, READ and KNOW is faster than to LIMP, BARK and STIR.
(3) Recall task
A researcher wants to discover which words are most easily recalled from a list. Subjects are asked to read a list and then tested on their recall of it. Their answers are checked in order to see whether they remember best:
- words from the beginning, middle or end
- longer words or short words
- words which are connected semantically to the words before them or words which are unconnected
(source: Field, John 2003. Psycholinguistics. A Resource Book for Students. Routledge)
Exercises: Elicitation of Responses