Temperature and bathroom scales

June 8, 2008 8:15 AM

Does temperature (ambient) affect the reading of an analog bathroom scale?

The reading I'm getting on my scale just dropped by 3 or 4lb, coinciding with a sharp change in air temperature (from 60F to 90F). I know these things are not perfect to begin with, so I could just chalk it up to the scale being quirky. (One day this winter, my weight randomly shot up by 3 or 4lb, and stayed there.)

My level of hydration hasn't changed, so I'm not blaming water (even though it seems a reasonable explanation if hot weather = sweat). I try to maintain a steady water intake.

posted by knave to Science & Nature (8 answers total)

My untutored instincts tell me "just chalk it up to the scale being quirky," rather than constructing a temperature-sensitive explanation -- or, at the very least, test the hypothesis with your scale during the next significant temperature change (in, say, 21550 A.D.).

Then again, I think this comment is quite helpful, which judgment is probably affected by the oppressive heat here.
posted by Clyde Mnestra at 8:21 AM on June 8, 2008

I would say it does affect the reading. Things expand and contract based on the temperature, and cheap bathroom scales don't exactly have high manufacturing standards. Hooke's Law (force on spring = constant x deflection) isn't affected by temperature, so my best guess without tearing open the scale is that there's some wiggle room at one (or both) ends of the spring inside the scale. Temperature rises, that wiggle room opens up more, basically increasing the size of the "dead" area around the zero point. Temperature drops, things inside tighten up, and the dial gets moving with less force applied.

I personally would only trust any bathroom scale to have a resolution of about +/- 5 pounds. If it tells you you gained or lost anything less, chalk it up to the margin of error on the scale.
posted by backseatpilot at 8:48 AM on June 8, 2008

It wouldn't move that much with temperature. Is there a zeroing wheel on the scale? You might have moved that accidentally.
posted by weapons-grade pandemonium at 9:25 AM on June 8, 2008

You can test it out by putting it in the freezer and measuring a known weight every hour or so.
posted by stavrogin at 9:29 AM on June 8, 2008

Aside from the many human factors (bladder and stomach/bowel contents, hydration, clothing etc.), I've noticed that cheaper bathroom scales can be quite sensitive to unevenness in the floor and the position of your feet on the scale. One digital scale I owned appeared to just pick any random number within 8 pounds of my actual weight.

My gf, who works as a diet counsellor, recommends buying a scale that has been approved for medical use. They don't necessarily cost much more, but they tend to be better constructed, properly calibrated and less sensitive to random environmental factors.
posted by le morte de bea arthur at 11:17 AM on June 8, 2008

That is a fairly big temperature change, and only a small variation in reading, so guessing that sensitivity to temperature is playing a role is very reasonable.

A good experiment is probably always best, so I'm with odinsdream/stavrogin on this one. Try a calibration measurement, and try that again at another temperature. If you do put it in the freezer, first put it in a clear plastic bag. Keep it in the bag even after you take it out of the freezer, to avoid condensation causing problems in your experiment (unlikely) or long term rust (likely). A properly arranged bag won't effect the measurement significantly.

The 'keep fat ass in chair' side of my personality tends to favour easily accessed reference material, but it has been a surprisingly hard thing to google. Terms like temperature, variation, spring, steel, scale, calibrate, hook's, and constant aren't leading to a good reference. There are lots of mentions of the idea in various contexts though.
posted by Chuckles at 12:37 PM on June 8, 2008

If it did it would be obvious because it would no longer be zero'd. My experience is that home spring scales are pretty bad regardless of brand. Its better to weigh yourself daily and do a weekly average to determine your weight.
posted by damn dirty ape at 12:40 PM on June 8, 2008

Response by poster: Well, I didn't get a chance to do any controlled experiments, but at this point, I think they'd be useless. The variation on the scale is simply too high to get any kind of precision.

Since I posted the question, it's only gotten worse, regardless of temperature. Today I stepped on it four times, and got four numbers ranging from 195 to 215. I think it's time to get rid of it.
posted by knave at 11:43 PM on June 25, 2008