ABSTRACT
I’d like to begin with a simple list of steps for constructing a research study that will guarantee statistically significant results when comparing two populations which differ linguistically. The procedures will produce results regardless of the subject of the research as long as language is in some way involved. But don’t worry about that. It always is involved.
Step 1. Choose a control group as much like you as possible. If studying children, your own or those of your neighbors and relatives will do best. The experimental group may be any other. Labels are unneeded but a wide range is available.
Step 2. Assume that your own dialect is standard and correct and all others are corruptions of your own!
Step 3. Encode all directions, questions, conversations with subjects, and statements in your own dialect. Again, an easy task since you know it so well.
Step 4. Judge responses as correct only if they are properly stated in your dialect.
Step 5. Use experiences, referents, and concepts which are drawn from the control group; that is, your own circle. A simple rule of thumb: if it is important to you, it is important.
Step 6. Create a pleasant experimental setting in which you feel comfortable. Surround yourself and your subjects with the tools of your trade, books, electronic equipment, etc.
Step 7. Provide a warm, friendly tester who speaks your dialect well and who can tolerate the odd characteristics of your subjects without reacting too extremely.
Step 8. Be sure to judge all data as deviation from the control group. A single hierarchical scale must be adhered to at all times. Remember you are the norm. All else is deficiency.
