As a scientist, I’m used to investigating things quantitatively. I try to answer questions with numbers and facts – with evidence, and this scientific way of thinking bleeds over into my everyday life. For example, when I noticed that the house seemed to feel colder than the thermostat claimed it was, I didn’t just trust that what I thought was true. Instead, I tried to answer the question quantitatively.
In this case, I compared the temperature displayed on the thermostat with the temperature reported by a kitchen thermometer I have for baking bread. And because good science requires multiple samples for good statistics, I didn’t just look at the temperature once and decide that I knew the answer. Instead, I compared the temperatures many times over the period of several days, at different times of day, and at different temperatures in the house. The result of this informal study seemed to indicate that the temperature difference is around 4 degrees.
While it seems likely to me that the 4 degree offset I’ve measured in this very informal “experiment” is probably approximately correct, the result is really not scientifically meaningful because I haven’t done things rigorously enough. To make my measurements truly scientific, I would need to properly examine the possible sources of error in my experimental system. Here’s a brief list of improvements I could have made to my simple experiment which would allow me to have more confidence in my answer.
- I should place the kitchen thermometer next to the thermostat. Because I left the thermometer in the pantry (where it lives, since it is used in the kitchen) instead of putting it in the hallway near the thermostat, it’s possible that I am simply measuring a 4 degree temperature gradient between the hallway and the pantry which actually exists. My own experience tells me that this is unlikely, since I would almost certainly notice a 4 degree difference in temperature as I move about the house. However, good science requires that I measure the temperatures in the same place for proper comparison.
- I have not determined that the kitchen thermostat is reporting an accurate temperature. This is important, since I have no scientific evidence that the thermometer is any more accurate than the thermostat. There are arguably reasons why the thermometer is more likely to be accurate than the thermostat, primarily that I bought it recently and so far it doesn’t seem to have led me astray. However, good science requires that I measure known temperatures with my thermometer to determine its accuracy. Boiling water and ice water would be good candidates here, since their temperatures (212F and 32F, respectively) are known and they can be easily made in my kitchen.
- I have not quantified the uncertainty in my measurements. The thermometer is analog and reports temperature on a dial with tick marks every 2 degrees. This means that at best I can only determine the temperature on the thermometer plus or minus 1 degree (i.e. I know it is a temperature greater than 62F but less than 64F, which should be reported as 63F plus or minus one degree). In fact, this error is even worse, since viewing angle has some effect on the apparent position of the needle on the temperature dial. So when I say the temperature difference between the thermometer and the thermostat is roughly 4 degrees, I’m not acknowledging that a better way to say this is that the temperature difference is between 3 and 5 degrees.
In the spirit of scientific rigor, I took my own advice and tested the accuracy of the thermometer, since this seemed the most likely source of significant error. In a small pot of boiling water, the thermometer read 211F (with error 1 degree). So I can conclude that the thermometer is accurate (to within error) as expected. To make this even more certain I could have tested the ice water, too. But even with just the one data point, I am quite confident in saying not just that the thermostat reports temperatures that are 4 degrees warmer than the thermometer, but also that the thermostat is in fact reporting a temperature that is 4 degrees warmer than the actual temperature in the house.
This sort of reasoning makes up what is known as the scientific method, and represents a way of investigating the world using quantitative evidence to discover reproducible facts about the world around us. Which is, at the most fundamental level, exactly what science is.
In this blog I will be attempting to share some of the science that we use every day, often without realizing it. I’ll explain such things as the chemistry of cooking, the biology of bread, the thermodynamics of air conditioning, and investigate simple experiments that can be done at home to answer age old questions: What’s the best way to remove those pesky red wine stains? What odors is baking soda most effective at neutralizing and why? And how clean is my counter if I clean it with vinegar instead of with bleach?