The Black and White Logical Fallacy and Surveys

The Black and White logical fallacy has a simple form: the respondent may only choose between A or B, while knowing that other possibilities exist. But this fallacy has several aliases: The false dilemma, the bogus dilemma, either/or, black AND white, and bifurcation.

This is a disjunctive fallacy, where the problem is in the false proposition that if  A is false, then B is true. There is no consideration that C, D…or so on could be true, and A or B could both be false, or even that A and B could both be true.

The black and white fallacy looks for affirmation of one premise and denial of the opposing or other premise or premises. In one aspect, alternative premises are eliminated.

In surveys, this presents a serious flaw in non experimental design. Surveys are generally non experiments because there is virtually no control over randomness of sampling from the population, there are biases in the population that voluntarily responds, or there is bias or falseness in the responses to the survey.

When mainstream media, for example, asks their viewers or listeners to “respond to our poll”, the only responses are going to come from viewers and listeners who just happened to catch that announcement, respond on line or make a phone call to respond. Worse, the question that is constructed with a black or white question does not allow the responses of “don’t know”, or  “other”. Finally, undesired responses can easily be removed from the final tally. Who would know?

In another construction:

Either P or Q

If P, then R

If Q, then R

Therefore, R!

This form  is designed to get approval for R, by making any survey answer one which results in support for R.

In some cases, the black or white question reveals information about the intentions, desires or even included a sales pitch by the surveyer. Revealing any information about the hypothesis or the desired outcome is to be avoided in surveys, or the results can be skewed, making the data worthless. A survey that seeks approval for either fresh or frozen food would be constructed as follows:

Which is nutritious: Frozen food or Fresh food?  This implies that if Fresh food is nutritious, then frozen food is not nutritious.

A better question: Which is more nutritious: Frozen broccoli or fresh broccoli? This question allows that BOTH fresh and frozen broccoli are nutritious, but one might be more than the other.

The even better survey would then compare perceptions of the nutrition in other frozen versus fresh foods, not just broccoli.

In the worst of political surveys, the respondents would be asked: Which would you rather be: Against the Patriot Act, or a patriot?  This question implies that if you are against the Patriot Act, then you cannot be a patriot. In other bad political questions: a choice of supporting immigration or being a bigot does not even begin to address the complexities of immigration issues. 

As a result, black and white logical fallacy in the social science non experimental survey can give information about the desired outcome, which can cause respondents to skew their answers. This fallacy eliminates alternative possibilities, or eliminates the possibility of testing for a null hypothesis. As a result, there is no internal or external validity to the experiment, and the data that is collected is easily proved to be unacceptable evidence in support of a hypothesis.