Originally Posted by SARDG
It may NOT vary during that test. The website article I referenced above claims that circuits built into cheaper scales can mask inconsistent measurements. If a 10gr weight is placed on the scale and actually measures 10gr, the circuits will take subsequent readings within a 'range' (as an example, 9.8-10.2gr) and display them as 10gr. The author suggests "cleansing the palette" of the scale between test weights by weighing a much heavier or lighter weight in between test weights.
These algorithms are allegedly how low-cost scales can use resolution to mask inaccuracy and imprecision.
It seems irresponsible for a scale that people depend on for safety to mask inaccuracy, but it doesn't surprise me, I know for a fact that my cheap bathroom digital scale does this. I can weigh myself, get off the scale, pick up a 5# weight and get back on the scale. It will read the same weight. If I weigh myself, take a shower, and re-weigh, I find I can gain or loose 5# just from bathing. I guess the time in the shower allows the scale to reset.
At any rate, I have said many times I don't trust cheap digital scales. You have no idea what is going on inside them. They could read correctly 99 times out of 100 and how would you know which ones are wrong? People pride themselves on the precision of their digital scales when as you pointed out, what they are really looking at is how many significant figures are being displayed.
For $22.00, at least the Lee scale has to be correct.