A common source of error in recorded text testing is the quality of the questions asked about the text. The quality revolves chiefly around the focus of each question, its translation, and its recording. Click on the links below to learn more.
Identify Semantic Categories
The easiest questions to form about a text will often revolve around only a few grammatical features. Such questions may miss the features that differentiate the two speech varieties that you are testing. Asking questions that focus on a variety of semantic categories and grammatical features will broaden the test and increase the likelihood of discovering barriers to comprehension. Click here to see an example.
Here are some no-no topics.
Clear questions that aim at a specific answer will be the best means of gauging comprehension. Though most poorly formed questions will be filtered out during pilot testing, any that remain will introduce confusion to the results. Click here to see an example.
Translating a question correctly into the subjects' language helps ensure that their answers reflect comprehension of the text and not the question. Checking the translation helps ensure that the meaning of your intended question is retained.