Today in class, we went over sections
5.5 Estimation and 5.6 Randomized Comparative Experiments. We learned that experiments study the response to a stimulus, to see how one variable affects another when we change existing conditions.
The two different studies we learned about were observational studies and experiment studies. Observational studies observe individuals and measure variables of interest but do not attempt to influence the
responses. The purpose of observational studies are to describe some group or situation, for example, a sample survey.
Experiment studies deliberately impose treatment on individuals in order to observe
their responses. The purpose of experiments are to study whether the treatment causes a change in the response.
Types of Experiments:
An uncontrolled experiment is an experiment where different variables may impact the experiment. For example, suppose two different groups of people, people who took online SAT classes rather than in-class SAT classes, and the experiment concluded that the people who took the online SAT classes did better on the SAT than those who took the in-class SAT classes. This is an uncontrolled experiment because many factors could impact the outcome of t
his experiment, for example, the people who took
online SAT classes could have been older, more experienced people who may have already taken the SAT, and the people who took the in-class SAT classes could have been younger people in high school. This would be a bias experiment, and not accurate.
A Randomized Comparative Experiment strives to get rid of the bias by randomizing which stimulus to apply to certain groups. For example, say we randomly took all of the people who intended to take the SAT classes (online and in-class) and mixed them up into the two classes, without knowing whether they would have originally taken the online classes or the in-class classes. This would make the experiment more accurate, because there would be no bias to this.
A control group in an experiment is the group that nothing is done to. The control is left alone so that the other variables may be compared to it. The control is the variable that is usually the "normal" result and that which you are testing against. For example, say we were testing whether the temperature has an effect on the breaking point of rubber bands. We took one rubber band and put it in the freezer, another rubber band and heated it, and left another rubber band at room temperature. The room temperature rubber band would be t
he control in the experiment, because it is the one that was not affected by any variables. This is also another example of an uncontrolled experiment, because different factors could have affected the out come of this experiment.
The last experiment we talked about was a Double Blind Experiment. This is an experiment where one does not know the difference between the stimulus and the control. For example, say a double-blind experiment was performed at a barbecue,
and Bob is trying to see how many people prefer Texas Pete over his own home made hot sauce. He gets his friend Billy to randomly put the hot sauces in two different bowls, labeling them A and B. Bob then takes the bowls and asks people which they prefer, and records the results. At the end of the experiment, Billy
tells Bob which hot sauce went in which bowl, and people ended up liking Texas Pete better. Poor Bob. This was a double blind experiment because Bob did not know which hot sauce was his, therefore he could not persuade the people in any way.
A good explanation of a double blind experiment is in the video below which we watched today, called "Scientific Method: How Double Blind Clinical Trials Are Done"
Next scribe is Farlz :) (Madison)
y than his peers.
