Experimental designs
In environmental science, we often have to choose between two main study designs: one where we take lots of samples from lots of places all at once – but only once twice (a survey) and one where we take lots of samples from only a couple of sites, on many occasions or for a long period of time (often called trend analysis). Describe an instance (not one we've already talked about!) when you think each approach makes more sense. How might choosing one design over the other affect your results or interpretation?
5 Comments:
Well I think that the one where you take one sample of more places is better when you are thinking about how location is a factor. The other one, when you take a lot of samples of only a few areas, would be more helpful if you're thinking about variables in specific environments, or how doing something to one of your samples makes it different from the ones that are like it. I dunno. It depends on what your question is I guess.
By Nick Jimenez, At January 11, 2007 7:33 PM
A large scale survey where you only check once or twice woud be something where you are only concerned about the before and after, like a diet. You take samples of all the insecure people before the try to lose whatever they think makes them look huge, and then after when all their insecurities have been doubled by the fear of a relapse into the McDonald's eating Ice cream scarfing ways of the past.
When you would take a large amount of samples from a smaller amount of sites would be something like when you are dealing with the effects of anti depressants on Emo kids. Then day after day you can observe them meander through their lives where they think that the entire world is out to get them and then after a while they realize that the world really doesnt care about them because they are emo and like to creat problems for their own personal endeavors.
That is experimental design.
And it rocks my socks.
Please excuse any misspelligns.
By Jordan Dolan, At January 11, 2007 7:46 PM
You could use a trend analysis if you wanted to see the effects of age or time on something, like if you were studying how listening to loud music through ear phones affects people's hearing over time. You wouldn't want to use a survey for this type of thing because different people who are different ages now have listened to music through ear phones differently; i.e., people who are older probably haven't done it as much, and people who are younger have probably done it more.
You could use a survey to study something on which time has little or no bearing. For example, you could conduct a survey to study how people living in different environments have different allergies to different things to see if the two things are correllated.
By Kaitlin Stenberg, At January 12, 2007 4:01 PM
So there are two studies types, long term short actual testing and long studying short term. Trend study make themselves best for medicine things like testing a new drug(not sure if that example was used) Trend studies would also be effect if you want to know the thing in question very well like the climate(i know this was used) and the effects of acid rain and other studies like that. So a trend study would be .... the effects of a treatment to acid rain by burning tums. A short term study would be something that takes little time but lots of trials. So what we are doing is that method, what else could be used mmmmm.... seeing if the more fast food in the place the fatter the people. This is perfect for my explaination of why one is better to use in this time then the other. A trend study would be if you did a before and after of building such a place and you would look at a long time this would give you a over time result and the one I mentioned this for would give you a number to connect the two. Sorry for the bad grammar, and bad spelling
By Ian Macomber, At January 14, 2007 8:09 PM
Kate identified a couple of nice examples - one for a survey and one for a long-term study. I think you've all got the concept.
By Sarah Nelson, At January 14, 2007 8:19 PM
Post a Comment
<< Home