Many studies rely on self-reporting for their data, but this can be a serious weakness. Even for such measurable items as height and weight, self-report can differ significantly from reality.
Brener et al (2003) compared the self-reported height and weight of over 2000 high school students to their measured height and weight, and found that on average, the teens reported themselves as 2.7 inches taller than they were and 3.5 pounds lighter. Female students were more likely to underreport their weight than males, but in a later study of high school students,
Brener et al (2004) found that female students were more likely to consider themselves overweight.
Spencer et al (2002) compared reported and measured values in a middle-aged British population, and again found height overreported and weight underreported.
Gillum and Sempos (2005) suggest that there is also a cultural aspect to this phenomenon; they found that the difference between reported and actual values was greatest in Mexican Americans. The point of concern in all of these studies is that the unrealistic reports suggest that the populations will not engage in weight-control practices, because they do not consider themselves to be overweight.
My driver's license says that I am 5'5" tall. That's purely wishful thinking. In reality, I'm probably about 5'4" (I think I'm a However, the fact that I overreported it doesn't mean that I don't know. When I 'm comparing myself to height and weight charts or talking to a doctor, I always look at the worst-case scenario: the shortest I think I could reasonably be measured and the heaviest I've been recently. However, for friends, my response varies. In short, I lie. However, I have to wonder about the standards these studies are using for "overweight" and "risk of overweight" In Brener 2004, for example, they report that 22.3% of students perceived themselves as overweight, compared to 47.4% who were "overweight or at risk for overweight." I have two problems with that assertion. First, only 26% of the students were actually in the "overweight category. They didn't give the students a "risk of overweight" category for self-assessment. Secondly, when half the students are "normal" and half the students are "overweight" (only 1.5% were in the underweight category by measured values), I don't know what basis for comparison they expect the students to use.