By Steven Pro
I am going to take a slight departure from my regular Impressions column this month. Instead of merely describing what I think about a particular type of equipment and the brands/models available, I will instead be detailing some product testing that I conducted. While I believe my standard Impressions column is worthwhile, it is inherently based upon anecdote and experience. In some instances, like power filters for example, I have handled hundreds of each brand and model. But with protein skimmers, in some cases I have only used three or four of some models. This limited experience can shade my opinion because of the flaws or alternatively fine performance of those units in that small sample. If I have only used a few of a product and all those were fine, it would give me the opinion that this brand was reliable whether or not that was the case. And of course, the opposite could be true. I could use a few of one brand and discover those were defective for some reason, but that may not necessarily be representative of the entire line.
While it is a commonly held belief that the standard box-style, swing-arm hydrometers are inaccurate, like so much else in this hobby, this belief is based solely on anecdotal evidence and experience and not based upon scientific study. After the release of the new and improved Marineland Labs/Aquarium Systems hydrometer, I decided to change that. I wanted to put this product through a series of tests to evaluate its accuracy and consistency when brand new, to reevaluate both after extended use and in some cases, a bit of abuse. I was particularly curious if the product was inaccurate and imprecise as manufactured or if they became inaccurate and/or imprecise only after use.
The first thing I should probably do is define some terms. (The following terms and definitions come from Beckwith, Buck, & Marangoni, 1982.) There is the true or actual value of something. In the strictest of senses, this value can never really be determined. It can only be approximated. There is also accuracy, which is the maximum amount by which the result is different from the true value. And finally, there is precision. Precision is how close the results are to each other. An extreme example of precision versus accuracy was given in the text book for illustration. “If all clocks in a jewelry store are set at 8:20 but are not running, the indicated values show precision but are accurate only twice per day". So, an instrument (or group of instruments) can be precise, consistently measuring the same value, but not necessarily accurate if those values are always incorrect. Additionally, an instrument can be occasionally accurate, but not necessarily precise if it gives wildly varying results. In the worst case, an instrument can be both wildly inaccurate and inconsistent, therefore neither accurate nor precise. Ideally, one would want an accurate and precise instrument, but could settle for a precise one if the error (the actual difference between the true and measured result) was known.
The design of the Aquarium Systems hydrometer has changed over the years, though the basic design is the same. The most recent design (far right) is the one tested for this article. Photo by Steven Pro