How can I test this hypothesis?

Question

Hypothesis: If the rusting of metal is related to rain, then exposing metal to rain will eventually cause the metal to rust.

What is a way that I could test this hypothesis?

P.S. Do you think this is a good hypothesis?

Thanks for your help!

Answer 1

To test: I'm presuming your only testing iron and iron alloys since you use the term rust.

Procure samples of several types of steel/iron (perhaps with several types of protective coatings (paint, galvanizing, tin plate, etc. Leave one set outside and record amount of rain exposure (or set up a hose nozzle to spray a constant, known quantity of mist on the samples for specified times (would be more quantitative and reproducible). Keep another set of samples indoors in a dry place. Observe the samples each day and note any evidence of rusting (you might want to take a digital photo of the samples each day for a record).

Answer 2

Well, you would need a control and a way to test this. In order to test to see if the rain is the cause, you would need to keep a piece of metal in the same environmental condition (temperature, humidity, light exposure, etc.) as the test metal, but the test metal would have to be exposed to the rain.

This experiment might be more of a long term project because it is hard to conduct the experiment; you are at the mercy of mother nature. But if you have the time, give it a shot.

Answer 3

You use the z-test when the population variance is known. You use the t-test if all you have is the sample variance. If you had the population variance that would likely mean that you have data on the entire population and you wouldn't need to do any hypothesis testing since you would have the actual mean. The t-dist can be formed by dividing a standard normal by the sqrt(chi square divided by its df) (remember that if you square a standard normal N(0,1) you get a chi square with 1 df) so let Z= (Xbar - mu)*sqrt(n)/sigma V = Summation(Xi-Xbar)^2/sigma^2 when you do the division notice that the sigma's (pop standard deviation) cancel out and that is why we do not need to know the pop variance when using the t-test. Way more helpful than Z because of this.

Answer 4

That is good...sort of..you should say "water" rusts metal..but if you have to say rain that is a true statement.. The physical property's of rain and water are. almost identical..so all you have to do is find a rusted nail and say it was because of water ex posher..Which it was..if you have to do it you self and want to see the process at work..just put a nail in water..It will take about 10 weeks...I would recommend finding a rusted nail.