2.3 Questionnaire
As has already been mentioned above, great care is necessary when creating your stimulus design. Hoffmann gives his readers a good insight into the problems connected with this task and we would like to summarise the most important points in this paragraph in order to help the reader understand the structure of our questionnaire. For a more detailed explanation of stimulus design, we refer to Hoffmann's article (cf. Hoffmann, Chapter 5). For reasons already stated above, we tried to create our questionnaire according to the main principles recommended by Hoffmann without abiding by every single rule of experiment design he deems necessary.
Magnitude Estimation aims for the elicitation of introspective data, especially informants' acceptability judgements on a particular grammatical phenomenon, which “get as close to grammaticality as possible” (Hoffmann, Chapter 5). The difficulty lies in minimising possible confounding factors which could influence an informant's judgement. An example for a confounding factor is the influence of particular lexical sets. Thus, when studying a grammatical phenomenon, it is important to provide the informant with a variety of sentences which contain the grammatical phenomenon in question, but in different lexical realisations (cf. Hoffmann, Chapter 5). Hence, our experimental stimuli are represented by ten sentences using five different verbs which can either take the reflexive pronoun or the zero variant (cf. Table 1).
Table 1: Experimental Stimuli
1. The Thames empties into the North Sea. 2. The Avon empties itself into the North Sea. 3. She troubles herself with grieves that are past. 4. He troubles very little with affairs of state. 5. He qualified himself for the Championship. 6. They qualified for the Olympic Games. 7. He overworked in the rush of the holiday trade. 8. She overworked herself excessively. 9. He couldn't keep from speaking out. 10. She couldn't keep herself from eating too much. |
Furthermore, one “additional problem that can affect psycholinguistic experiments is the formation of implicit hypotheses” (Hoffmann, Chapter 5). In other words, an informant should not solely be exposed to experimental stimuli, because he or she might become aware of the experiment's aim and be affected accordingly (cf. Hoffmann, Chapter 5). In order to preclude this effect, linguists include grammatical and ungrammatical fillers in their questionnaires. If possible, their number should equal the number of experimental stimuli (cf. Hoffmann, Chapter 5). Again, we adopted this practice for our study and the result can be read in Table 2.
Table 2: Fillers
Grammatical fillers: 1. Alex looked at the mirror. 2. He himself said it was right. 3. He did it for himself. 4. They asked themselves whether to go or not. 5. I looked at myself in the mirror. Ungrammatical fillers: 1. She troubles myself with grieves. 2. Robin washed itself. 3. She couldn't keep from herself eating too much. 4. They helping themselves. 5. He talked himself to. |
Last but not least, “[f]atigue, boredom, and response strategies the informant may develop over the course of the experiment can have differing effects on sentences judged at various points in the entire procedure” (Cowart quoted in Hoffmann, Chapter 5). Hoffmann therefore recommends that researchers randomise the sentences before presenting them to their informants. The application of this rule to our experimental stimuli as well as the fillers results in the following sequence of sentences (cf. Table 3).
Table 3: Randomised sentences
1. She troubles myself with grieves. 2. Robin washed itself. 3. Alex looked at the mirror. 4. The Thames empties into the North Sea. 5. She couldn't keep from herself eating too much. 6. He himself said it was right. 7. He did it for himself. 8. They helping themselves. 9. They asked themselves whether to go or not. 10. She troubles herself with grieves that are past. 11. He talked himself to. 12. He qualified himself for the Championship. 13. He overworked in the rush of the holiday trade. 14. He couldn't keep from speaking out. 15. The Avon empties itself into the North Sea. 16. She couldn't keep herself from eating too much. 17. They qualified for the Olympic Games. 18. I looked at myself in the mirror. 19. She overworked herself excessively. 20. He troubles very little with affairs of state. |
Of course, the randomised sentences have to be complemented by the reference stimulus, which must be visible throughout the experiment. In the case of our study, the sentence Marc only thinks of himself was chosen to be the reference stimulus.
Created with the Personal Edition of HelpNDoc: What is a Help Authoring tool?