Select Final Questions

From SurveyWiki
Revision as of 12:06, 18 April 2011 by Katie Crystal (talk | contribs) (Formatting)
Jump to navigationJump to search
Recorded Text Testing
Background Research
Intelligibility Interviews
Choose Kind of Test
Choose Test Points
Materials
Obtain a Text
Develop Questions
Create Introduction Text
Assemble Pilot Test
Administer Pilot Test
Select Final Questions
Build Reference Tests
Develop Post-RTT Questions
Administer Text Test Set
Process Scores

For one reason or another, not all of the questions created for the pilot test will be good ones or be understood clearly by the subjects. Taking these poor questions out of the test will leave only the ones that were understood and answered correctly by everyone, which will make the instrument more reliable. If the speakers of the test language were able to answer every question, then you can reasonably expect any person who understands the text to provide the correct answers also.

Once standardization is applied to the pilot test scores, you need to mark for deletion the ones that will not be included in the final tests. When deleting question tracks from the pilot test to form the hometown test, it is possible to delete the wrong track or reassign track numbers through mismanagement. Practicing this procedure will help you keep all the right tracks and maintain a correct track table.

Open the following document to learn how to evaluate and choose final questions.

Suggestions for removing and deleting questions

These are the questions that should be eliminated from the final test:

  • Questions that were answered incorrectly
  • Questions that were answered incompletely – if you find that some subjects provide less detail in their answers than other subjects, this may be an indication that the question was not understood equally or that hometown subjects have differing ideas about what to include in the answer. You do not want to risk either situation in the final test.
  • Questions that were unclear to the pilot test subjects
  • Questions that confused the subjects
  • Questions that were answered only after a long delay – if subjects had to stop and think about the meaning of the question or the way in which they should respond, then you don't want to put final test subjects in the same position with potentially less comprehension.

You also need to remove from the test:

  • Questions that repeat a type of focus or semantic category (Casad 1974). Keep the question that creates the most even distribution of questions throughout the text.

Do not eliminate:

  • Any question answered correctly that has a unique focus on a particular component of the text.

In short, you want the largest possible inventory of questions that have never been answered incorrectly by the hometown population and which each have a unique focus.

When deleting questions tracks from the MiniDisc itself, remember to:

  • Create a new track change table – the new track table will be the guide for the insertion of questions in the speech of each test point.
  • Backup the disc that contains the whole pilot test – if you make a mistake, this backup could save you hours of work.
  • Note the changes in tracks – remember that the numbers of the tracks after the one deleted will move up.