Module 3: How To Analyze and Interpret Data

Question 0

Define:

Control (of an experiment)

Answer 0

The variable or part of the experiment to which all others will be compared

Question 1

Define:

Experimental variable

Answer 1

An aspect of an experiment that changes during the course of the experiment

Question 2

Define:

Blind experiments

Answer 2

Experiments in which the participants do not know whether or not they are part of the control group

Question 3

Define:

Double-blind experiments

Answer 3

Experiments in which neither the participants nor the people analyzing the results know who is in the control group

Question 4

When is an experimental variable good and when should it be reduced or eliminated?

Answer 4

An experimental variable is good when you are using it to learn something from the experiment. And experimental variable should be reduced or eliminated when it affects the results of the experiment but you do not learn anything from it.

Question 5

Questions 3 through 7 refer to the following story:

Answer 5

A consumer laboratory decides to test the effectiveness of different laundry detergents. Five white shirts are stained with grass things and put into five different washers. In one washer, no laundry detergent is used. And the other four washers, four different types of laundry detergent are used, one type in each washer. Water from the same source is used to fill each washer. The washers are then turned on for the same amount of time and same kind of cycle, and once they are finished, the shirts are examined by eye to see which is the cleanest.

Question 6

3. What is the control group for this experiment?

Answer 6

3. The control is the shirt that is being washed with no laundry detergent at all.

It is possible that all the detergents are so bad that they have no real effect on the cleanness of the shirts. The only way to tell is to compare it to a shirt that was washed and no detergent.

Question 7

4. What is the experimental variable that will be used to learn something from the experiment?

Answer 7

4. The experimental variable that can be used to learn something from the experiment is THE TYPE OF DETERGENT USED.

Question 8

5. What are the experimental variables that need to be reduced or eliminated?

Answer 8

5. There are at least four unwanted experimental variables. First, THE WASHERS ARE DIFFERENT. It is possible that some clean clothes better than others. This affects the results of the experiment, because you will not know whether the difference is cleanliness due to the washer or the detergent. In addition, THE WATER CAN BE DIFFERENT TEMPERATURES, which will affect the outcome. Also, THE SHIRTS ARE ALL DIFFERENT. Some fabrics are easier to clean than others. Finally, THE AMOUNT OF GRASS STAIN WILL BE DIFFERENT IN EACH SHIRT, because there is no way to stain shirts equally.

Question 9

6. What could be done to reduce or eliminate the unwanted experimental variables?

Answer 9

6. The experimental variable of the washers can be reduced by making sure our washers are the same brand and model, and by making sure they are all relatively new. This will reduce the differences among the washers. You can reduce the differences in water temperature by monitoring the temperature of the water as it enters the washer and making adjustments to keep the temperature the same. The experimental variable of the shirts can be reduced by making sure they are all from the same manufacturer, the same style, and the same fabric. That way, they are as close to identical as possible. Finally, the experimental variable of the amount of stain can be reduced by examining each stain carefully and trying to make sure they are all as identical as possible.

Question 10

7. Are the data collected objective or subjective?

Answer 10

7. The data being collected are subjective.

Think about it. Each person’s definition of “clean” is different. Also, the shirts are being examined by eye. This makes it hard to say exactly how much stain is left on each shirt. If you could chemically examine each shirt and determine precisely how much grass stain was left after washing, you would have an objective measurement. However, to have someone just look at a shirt and decide whether or not it is cleaner than another shirt is subjective.

Question 11

Why can a carefully placed needle float on water, even though a needle is denser than water?

Answer 11

The needle floats because of surface tension.

Question 12

What does soap do to the surface tension of water?

Answer 12

Soap reduces the surface tension of water.

Question 13

A student tries to float a needle on water. He succeeds, but only after several attempts. He then tries to float the same needle on another liquid. Although the needle sinks when dropped in the liquid, it is much easier to lay the needle on the surface of the liquid and make it float than it was to get the needle to float on water.

Compare the surface tension of this liquid to the surface tension of the water.

Answer 13

The liquid must have a larger surface tension than water.

This is because the needle floats more easily on the liquid than it does on water.

Question 14

A new, fat-free potato chip comes out on the market. A few months later, there are reports that some people get severe stomach cramps a few hours after eating the chips. You must do an experiment to see if the chips cause severe stomach cramps. A group of people volunteer for the study. Describe how you would design the experiment. Also, indicate whether the experiment should be single-blind, double-blind, or neither.

Answer 14

You should give half the volunteers the fat-free potato chips and the other half should get potato chips that have been on the market for years and seem to have no problems associated with them. The latter group is the control. The volunteers then can keep a log or you could observe them for the next few hours to see if any stomach cramps occur. If more cramps occur in the group that ate the fat free chips than what occurred in the control group, the allegations could be true. This should definitely be a double-blind experiment. Comparing how two groups of people feel after eating is subjective. There’s no way to get hard numbers from such a study. Thus, the person analyzing the data needs to be blind as well.

Question 15

A study is done to see if a certain herb can increase a student’s concentration skills. A group of students volunteer for the experiment, which consists of giving the students a pill made of either sugar or the herb. The students then take a series of math test. The test scores of the students who take the herb will be compared to those who took the sugar pills. If there is a difference between the average test scores of the groups, it could very well be the result of the herb. Should this be a single-blind experiment or a double-blind experiment or neither?

Answer 15

This should be a single-blind experiment.

If the students know whether or not they were given the herb, it might influence how they take the test. However, since the data being collected are objective (measurable numbers), there’s no reason for the person analyzing the data to not know who is in the control group and who isn’t. Of course, if the person wants to avoid any appearance of dishonesty, the experiment could be done as a double-blind experiment.

Question 16

A farmer has four different cornfields. The government pays him to experiment with three different kinds of herbicide (weed killers). He sprays nothing on one field, and he sprays each of the other fields with a different brand of herbicide. He then records the numbers of tons of corn he produces from each field, and reports back to the government about which herbicide produces the best crop. Should this be a single-blind experiment, a double-blind experiment or neither?

Answer 16

The study should be neither single-blind nor double-blind.

The experimental subjects are plants. They cannot know whether or not they are a part of the control group or not. Thus, the whole idea of a single-blind experiment is kind of irrelevant. Also the data being collected (the weights of the corn) are objective. The farmer can’t bias the result, so there is no need for the farmer to be blind.

Question 17

A researcher is trying to determine if there are any differences between how homeschooled students play as compared to how publicly school students play. She wants to get a group of publicly school students and a group of homeschooled students together and observe how the children play with one another. She will record her observations and then try to see if the ways in which homeschooled children play with each other are different from the ways in which publicly school children play with each other. Should this be a single-blind experiment a double-blind experiment or neither?

Answer 17

This should be a double-blind experiment.

If the students know whether or not they are in the control group, it might influence how they behave. For example, the homeschooled students might be on their best behavior so as to give the researcher good result for homeschoolers. In the same way, the researcher’s observations are subjective. There is no way to precisely measure how one child plays with another. It will depend heavily on the researcher’s preconceived notion. Thus, the researcher must be as unbiased as possible and should therefore not know who is in the control group and who is not.

Question 18

A scientist comes up with a revolutionary drug that he thinks will allow people to lose weight without dieting. They simply need to take the pill and their excess weight will slowly disappear. He decides to test the pill by getting two groups of volunteers together. The first group will take a fake pill, the second group will take his new drug. Each volunteer will be weighed once a week for 12 weeks. Should this be a single-blind experiment, a double-blind experiment, or neither? What results will indicate that the new drug is effective?

Answer 18

This should be a single-blind experiment.

The subjects need to be blind because telling the person that he or she is on the “real” drug might affect his or her behavior. Thus, everyone must be blind as to whether or not they are in the control group. However, there is no need for the researcher to be blind, because the data are completely objective. The data are a series of numbers that cannot be affected by the researchers preconceived notion. Thus, whether or not he knows who is in the control group cannot affect the outcome of the experiment. (This experiment could be performed as a double-blind study. That way the researcher could not be accused of altering the measurements.)

The data will indicate whether the drug is successful or not.