Activity 1-1

How many psychiatrists thought that people would go higher than 13 (very strong shock; 195 volts) on the scale?

1

Milgram also asked college students to predict the maximum number of volts people would administer. Create a frequency distribution table and graph for the variable CollegeStudentPredictions. How many college students thought that people would go higher than 13 (very strong shock; 195 volts) on the scale?

1

The predictions from the psychiatrists spanned a wide range of values. What was the most common response?

10

What is the psychiatrist prediction at the 87.2th percentile? In other words, what is the predicted "administered shock" value that has 87.2% of the distribution equal to or lower than it?

10

The highest level that one college student predicted that people would go to was level 14. What percent of people went higher than level 14 in Experiment 1?

100.00%

What percent of psychiatrists thought that people would go higher than 13 (very strong shock; 195 volts) on the scale? If your software gives a percent as well as a valid percent column, be sure to use the valid percent.

2.56%

What was the highest level of "administered shock" that a psychiatrist predicted?

20

What percent of college students thought that people would go higher than 13 (very strong shock; 195 volts) on the scale? If your software gives a percent as well as a valid percent column, be sure to use the valid percent.

3.23%

The highest level that one psychiatrist predicted that people would go to was level 20. How many people went higher than level 20 in Experiment 1?

35

Before beginning this study, Milgram asked a group of psychiatrists to predict the maximum number of volts people would administer. Download the Milgram data file on the textbook website and create a frequency distribution table and graph of the variable PsychiatristPredictions. How many psychiatrists made predictions?

39

The actual data from Milgram's first study are in the Milgram data file. Create a frequency distribution table and graph for the Experiment1 variable. How many people participated in the study?

40

In Experiments 1 and 2, the Learner and Teacher were in different rooms. In Experiment 3, the Learner and Teacher were in the same room. Create a frequency distribution table and graph for Experiment 3. What percent of people went all the way to level 30 in Experiment 3?

40%

What was the second most common response from psychiatrists?

5

In Experiment 1 there was no verbal communication from the Learner after the testing session began. In Experiment 2 (done with different participants), the Teacher could hear the Learner protest through the walls. The protests became increasingly vehement as the shocks increased until all sound ceased at 330 volts and the Learner stopped responding/pressing buttons during the memory test. Create a frequency distribution table and graph for Experiment 2. What percent of people went all the way to level 30 in Experiment 2?

62.5%

The numbers in the data file refer to the maximum shock level each person gave. Milgram defined "obedient" as going all the way to level 30 (450 volts, XXX). What percent of people went all the way to level 30 in Experiment 1?

65%

What percent of people went higher than level 20 in Experiment 1?

87.5%

In addition to learning from the legitimate ethical concerns raised by Milgram (1974), we should also learn the importance of

Both answers are important considerations when interpreting scientific results.

Teachers thought they were giving shocks by pressing levers that increased in steps of 15 volts. There were no possible responses between any two levers. Does this suggest that the dependent variable is discrete or continuous?

Discrete

Experiments 1 and 2 yielded similar numbers of people who were classified as "obedient" in that they delivered the maximum level of shocks. However, the two distributions of scores are not identical. For which experiment did some people stop shocking the learner sooner?

Experiment 2

In Experiment 1, Milgram recruited a sample of 40 men between the ages of 20 and 50 from Connecticut. If we use the results from this sample to infer what the results would be if he had collected data from the entire population of 20 to 50-year-old men, would the purpose of the study be descriptive or inferential?

Inferential

The dependent variable in this study is how far each Teacher went on the shock scale. A score of 0 would mean that they were unwilling to administer any shocks while a score of 30 would mean that they went to the very end of the scale (450 volts labeled XXX). The fact that each numerical label on the scale increased by exactly 15 volts suggests that this "shock administered" variable is measured on a ___________ scale of measurement.

Interval/Ratio

As shown in Figure 1, in addition to the numerical labels (e.g., 15 volts, 30 volts) the Teacher saw verbal labels that grouped the specific voltages into increasingly severe punishments (e.g., slight shock, moderate shock, strong shock, etc.). These verbal labels represent the ____________ scale of measurement. (Nominal, Ordinal, or Interval/Ratio)

Ordinal

Choose the best summary of the results from Experiments 1-3.

Participants were more obedient to authority than was expected. Many administered shocks at the highest levels. When participants were able to communicate with the learner or were in close proximity to the learner, they administered less severe shocks.

If the ___________ of participants represented the ____________ of men in Connecticut between the ages of 20 and 50 well, the study would have a relatively small amount of sampling error.

Sample, population

Look at the graph you created. Many statistical analysis programs exclude categories with no responses from being listed on the x-axis. In this study, the possible categories ranged from 0 to 30. In your program, are all categories from 0-30 listed on the x-axis or are some response categories omitted? This is important because graphs look very different if response categories are omitted from the x-axis.

Some response categories are omitted.

If the study has a relatively small amount of sampling error, the sample ________________ will be similar to the population _________________ .

Statistics, parameters