Wednesday, July 27, 2011

Knowledge and Belief

I was just directed to this announcement concerning an NSF survey on science literacy. The bottom line is that the NSF is deciding to change the wording of two questions in the survey. The original wording is “Human beings, as we know them today, developed from earlier species of animals,” and “The universe began with a huge explosion.” . The new wording is “According to evolutionary theory, human beings, as we know them today, developed from earlier species of animals” and “According to astronomers, the universe began with a huge explosion.” (emphasis mine). It is noted that there will be a transition period with the questions, with half of the surveys containing the new questions and half the old, to determine its effect.

The stated goal for this change, from the NSF, is to separate knowledge from belief. You might believe that humans are created in their present form, 6000 years ago, but the new questions try to ascertain whether you know that “evolutionary theory” says something different. Is this an important distinction? Is this what we really want to measure? Which is more important for a society? What is the difference between knowledge and belief?

It is quite clear that there will be at least one effect for this rewording: given that the US falls way behind other countries on science literacy, especially with these particular questions, the rewording will most likely increase these numbers with no other work done.

Definitions and Concepts

Beliefs are representations of the world. Specifically, they are representations that we hold to be correct for the real world...as opposed to hopes, which are also representations of the world by not ones that we hold to be necessarily correct. Along with beliefs we always have a confidence in the belief, specified as a probability (either explicitly or implicitly). Knowledge is simply that collection of beliefs that we hold with such high probability or, in other words, with such confidence that we do not significantly doubt them. The belief that the sun rises in the east each morning is considered knowledge for the reason that we hold it with an extremely high probability. This is not just as a result of an inductive reasoning process (i.e. it always has, in our experience, risen in the east each morning) but because it is part of a larger body of knowledge (i.e. astrophysics) for which it is just one consequence within a whole host of other well-established predictions.

Now, on to scientific literacy. The NSF defines scientific literacy as “knowing basic facts and concepts about science and having an understanding of how science works.” Why is it important? Again, the NSF: “It is valuable not only in keeping up with important science-related issues, but also in evaluating and assessing the validity of any type of information and participating meaningfully in the political process.” The question we must ask is, does the new wording measure scientific literacy better than the old wording? To do this, we need to outline the four possible types of people answering the two forms of the questions:

  1. people who answer “yes” to the old and “yes” to the new
  2. people who answer “no” to the old and “no” to the new
  3. people who answer “no” to the old and “yes” to the new
  4. people who answer “yes” to the old and “no” to the new
The wording change doesn’t change cases 1 and 2, adds case 3 to the “yes” category and it introduces the erroneous case 4. The cases can be summarized in another way, like

  1. people who know both that, say, the universe began with a big explosion and that astronomers claim that this is true. This is indicative of scientific literacy.
  2. people who don’t know, or do not believe, that the universe began with a big explosion and that don’t know that astronomers claim that this is true. This is indicative of scientific illiteracy.
  3. people who don’t believe that the universe began with a big explosion but know that astronomers claim that this is true. (more on this below)
  4. people who know that the universe began with a big explosion, but do not believe that astronomers claim that this is true. This might at first seem to be a totally unreasonable and marginal case, but I think it is more significant than perhaps is generally appreciated. These people might think that the new wording is a trick question (e.g. they might think that physicists, as opposed to astronomers, claim that it is true). I’ve had students answer questions in this way, so it is not quite as uncommon as one might think. These students overthink the problem: they know the fact, but are distracted by the extra complexity of the question, thinking that the test is trying to trick them.
Case 3: The Religious Believer

The only reason these particular questions were modified was because of the prevalence of religious belief. How do we know this? We don’t see a proposal to change “The Earth orbits around the Sun and takes a year to do it” to “According to astronomers, the Earth orbits around the Sun and takes a year to do it.” Why? Because no religion (now) has a stake in the answer to that question, and thus have no objection to the claim. Of course, if you go back to the days of Copernicus this was a different story and people were severely punished for too strongly making such a claim. The two questions that are proposed to be changed in this way are precisely the two concepts that crop up in every creationist tract, and are clearly the two major stumbling blocks for a literalist reading of the Bible or the Quran.

Aside from the motivation for the change, we can ask the question whether it is accomplishing something important anyway. Are these Case 3 people, who would answer “no” to the old question but “yes” to the new question, demonstrating scientific literacy? I don’t think so. What they’ve confirmed is that they know that some scientists claim that the universe began with an explosion, but they don’t believe it. This means that they don’t accept the data, or the methods, or both. If the question were about something on the fringes of science, then perhaps this is fine, but it isn’t the case with these two questions. Evolution theory, for example, is as well established as the Round Earth theory and the Germ theory of disease. To deny it is to deny all of the independent work in molecular biology, embryology, ecology, etc... which support it. Even though they may know that fact that biologists support Evolution theory, they have not demonstrated any scientific literacy in terms of “evaluating and assessing the validity of any type of information and participating meaningfully in the political process.” The same can be said of the Big Bang theory, to a slightly lesser degree (i.e. there isn’t quite the volume of completely independent fields of study supporting it, as there is for Evolution, but the data is nearly incontrovertible anyway). To deny either idea is akin to denying the Germ theory of disease.

Bottom line

Imagine someone answering “no” to the question “The world is round” but answers “yes” to “According to geographers, the world is round”. Would they be demonstrating scientific literacy? I don’t think so. Do we want to pander to the religious-motivated ignorance in this country, for the sake of increasing the appearance of scientific literacy? I don’t think so.

2 comments:

  1. While this isn't a problem we should have to deal with, perhaps this pandering is a necessary evil if we want to get an accurate picture of what Americans really know, rather than what we choose to believe for religious reasons. More information is never a bad thing, and if we ask both whether or not people know that, according to evolutionary theory, and whether or not they personally agree with the statement, as was decided by the second NSF workshop, we will learn whether our problem lies with people not understanding evolution at all, or with people understanding evolution yet still refusing to believe in it. That said, a question like, "according to evolutionary theory, humans developed from earlier species of animals," does not really test respondents' knowledge of evolutionary theory. The best solution might be to ask a few, more challenging, questions about evolution, qualifiers and all, so we can determine whether our problem lies with lack of knowledge and understanding or refusal to believe in evolution despite having an understanding of it.

    This is basically a longer version of this comment, from my blog:http://rocketpropelledpeacocks.blogspot.com/2011/08/national-science-board-which-governs.html . Not trying to self promote, I just don't want to repeat myself more than I already have.

    ReplyDelete
  2. This was a very thoughtful comment, and I apologize for not getting to it sooner. I think the plan of the survey is to eventually *replace* the questions, not ask both. So I think if we had to choose between the two types of questions, I'd prefer the one that is closer to the way they would make choices rather than what they consider to be "what I was told in school". I think you're not just sacrificing "a few of our rigid ideals", as you say, when choosing the latter...you're getting rid of (perhaps the only) relevant information. I agree that more detailed questions would be better, but I think the length of the survey then becomes an obstacle to many people finishing it.

    Finally, I am disturbed by the fact that the *only* questions under consideration are the ones that have a religious impact. Does that bother you at all?

    ReplyDelete