In teaching and assessment a foil is simply an incorrect alternative. Any time a choice is given the foil itself can make or break a response’s accuracy. As an example, consider this picture:
Now, here are four questions designed to determine your knowledge of the picture’s subject.
1) What is this?
2) Is this uranium, pyroxite, or feldspar?
3) Is this plagioclastic-orthonograph feldspar or uranium?
4) Is this a type of fruit or uranium?
Much can be ascertained about one’s uranium knowledge depending upon which questions can or can’t be answered. We can learn that somebody that can answer the question without foils (labeling, in this case) knows his rocks. Conversely, when using bad foils nothing may be discovered at all. Most second graders could answer the fourth question correctly which, of course, tells more about the child’s knowledge of fruit than uranium. The third question’s foil is almost as bad. If someone answers “uranium,” how do you know it’s not simply because the foil was so hard to pronounce? While these examples may be extreme, they illustrate the significance that seemingly simple framing and foils can have on good assessment.
Three main foil factors can increase or decrease a response difficulty independent of a respondent’s knowledge of the target stimulus.
1) number of foils
2) semantic similarity of foils to target
3) syntactic complexity of foils
So, the greater the number, the closer the similarity, and the less complex the syntax, the greater the difficulty. Increasing difficulty may be a good thing, such as when testing more advanced knowledge of a subject. Decreasing difficulty can also be desired for lower functioning students.
For example, when measuring a respondent’s knowledge of labeling vowels, you may show three pictures and ask “Which one is a vowel?” If you show the letter O and two animals, the respondent requires very little (but still some) actual knowledge of vowels to point to the O. If you use one similar foil, and one semantically different foil, you learn more. A student that points to the picture of a tree rather than an O obviously knows less about vowels than a student that at least points to a consonant. The latter student at least knows that vowels are letters.
One final note about using choices – their use is also a good way of promoting labeling in teaching. A child that can’t label can often provide the answer to a choice. Incremental manipulation of the choices can be good therapy. One final example:
Instructor: “What is the coldest continent?”
Student: “I don’t know.”
Instructor: “Is it Antarctica, Asia, or Canada?”
Student: “Canada.” (Providing the final answer, even when far from correct, indicates a great deal about the student’s target knowledge.)
Instructor: “No. I’ll give you one more try. What is the coldest continent? Is it the United States or Antarctica?”
Hopefully the student will now correctly answer, and in doing so, not only will the instructor have gained assessment information, but the student will have actually labeled the correct answer with cues. This increases the likelihood of later labeling the answer with fewer cues. The only way that this can happen is if learning has also happened.