How do we test hypotheses?

Answer 1

by collecting data. Experimentation In our experiments we attempt to set up conditions so that we know that the outcome of the experiment is the result of the variable that we are manipulating. I hypothesize that a given substance will have a positive effect on plant growth. I will set up several experimental conditions, holding everything constant in each group except the substances given for growth. I observe the growth of the plant, and draw my conclusions. The hypothesis may hold, or it may need to be modified. For example, I notice that the substance that inspired the hypothesis may have caused growth that was too fast, leaving spindly and non-productive plants. I may want to continue the development of the substance, or perhaps find a way to blend it with other substances, and then continue the experimentation. _______________ Here is another example: I am investigating the properties of water, and I need to know all I can about how water, ice and heat interact. I know that when water gets cold enough, it freezes, and that when ice is exposed to heat, it melts. I might come up with a hypothesis-- a first attempt to understand what is happening with water and ice. I hypothesize: Since water is the liquid state of H2O, and liquids are warmer than their solid forms, water must be warmer than ice. It must be that as soon as water forms from melting ice, it is at least a little warmer than the ice. After all, at zero degrees water becomes ice. At this point, I do not know for sure, but I am speculating (hypothesizing) based on the general information that I have. I wonder how I can test my hypothesis: Let us assume that I can measure the temperature of ice and of water with great accuracy. I should be able to easily make a mixture of ice and water, while measuring the temperature of each independently.

• If my hypothesis is correct, I will find that at one atmosphere of pressure and in a chamber kept at exactly zero degrees Celsius the ice will warm to zero degrees Celsius (if it starts out colder than that) and it will get no warmer. When it is warmer, it is water and not ice.
• I will also observe that the water will be at least a little above zero degrees Celsius when it first melts away from the ice.
• I apply heat to the ice and some ice melts.

To my surprise and annoyance, I find that the water is exactly zero degrees Celsius. We have a problem. My fellow engineers need an answer, and I've been hyping my hypothesis all over the place..... Eureka! I have found it! Keeping the chamber at zero degrees is the problem; as soon as the ice melts, the chamber air starts to cool the water, and it must also be cooling the thermometers, leading to an experimental error in the readings! I'll keep the chamber at 5 degrees Celsius.... I go back to the lab and I notice that my containers now all have a mix of ice and water, and the chamber is at exactly zero degrees Celsius. Not only that, all the ice is now at zero degrees Celsius, and so is the water! Exactly! Now I am just plain confused. I leave the chamber exactly as it is for several days and nothing changes. Everything in the chamber is verified to be at zero degrees Celsius, water is not in the process of freezing, and the ice is no longer melting. It just all stays in equilibrium. If my hypothesis had been correct, there would be no water in the chamber, because supposedly at zero degrees Celsius all water is ice, and "water is [so I hypothesized] always at least a little above zero degrees". My hypothesis has tanked; time for a new one. I wonder and wonder, and spend a few sleepless nights. Then I realize-- it has something to do with heat; it must. It has something to do with adding or with taking away heat... Hmmm. Testing a hypothesis is something like that. You wonder, and you explore. Hypotheses develop, and may become well-known theories when they seem to be able to withstand test after test designed to expose them false. When you think about it, how would you develop an experiment that exposes a hypothesis as true? They would be the very same tests, with results that support the hypothesis. Reality determines the outcome, not really the 'test'. A silly or inconsequential test will still show results in support of reality. But as mentioned above, even if all completed tests support a hypothesis, that is not to say that some unexpected approach will not be applied tomorrow.