A Thesis on Effects of an Authority Figure on Obedience

Abstract

We conducted this study to test if findings from previous studies of obedience would generalize to college students working in a lab. We collected data from a sample population of 50 college students interested in psychological research. We randomly split the participants into two groups, one group worked with a “strict” experimenter, and the other worked with a “relaxed” one. To measure obedience, we recorded how many rules each participant broke while collecting data during an observation experiment. We found that there was no significant relationship between the group placement and the amount of rules broken.

We can conclude from these results that while there may be some pattern to rule-breaking behavior, the type of authority figure has no effect on obedience; and that results from previous studies do not generalize to college students working in research.

In 1963, a Psychologist named Stanley Milgram published his study on blind obedience to an authority figure, now infamously known as “The Milgram Experiment.” Milgram found that the majority of the men who participated willingly shocked a patient whom they could not see with increasingly higher voltage, because they were ordered to do so by an experimenter.

His experiment became one of the most famous studies of obedience in history due to its shocking discoveries (Blass, 2004). Despite this, the study left many important questions unanswered, such as, why and when are people obedient? What makes someone an authority figure’? What real life applications does this information have? Many researchers have tried to find answers to such questions by conducting their own studies,  often taking inspiration from Milgram’s work.

Get quality help now

Bella Hamilton

Verified

Proficient in: Human Nature

5 (234)

“ Very organized ,I enjoyed and Loved every bit of our professional interaction ”

+84 relevant experts are online

Hire writer

The present experiment is one such study. We seek to understand how the perceived “strictness” of an authority figure may affect the rule following behavior of a college student working with a professor; specifically, in a research setting. For the purpose of this experiment, obedience can be defined as a social phenomenon in which a person follows instructions or orders given to them by someone with authority.(Colman, 2009). A person practices obedience each time they follow a rule given to them by a parent, teacher, boss, or other figure of authority; obedience is the act of following the rules.However, rules are not always obeyed. One may know from personal experience a rule or law that is frequently broken, such as, “don’t eat too many sweets,” or, “don’t drive over the speed limit.” Although these rules exist for good reason, (eating too many sweets may negatively affect your health, and driving faster than the speed limit increases your risk of an accident), they are still broken quite often. Why?

There are many factors that influence why a rule is followed or broken. One of these factors is the perceived “legitimacy” of the authority behind the rule. What does one mean when they describe an authority figure as “legitimate”? In the field of psychology, legitimacy is conceptually defined as a characteristic that a person may have which causes others to see their authority as justified (Tyler, 2006a and Tyler, 2006b). When someone’s authority is seen as legitimate, others are more likely to follow orders and be obedient to their authority, even when that person isn’t there to make sure their rules are followed (Sunshine and Tyler, 2003). Furthermore, they don’t have to have instrumental control over the rewards and punishments following a behavior in order to control it, meaning they don’t have to waste as much time and resources to ensure their rules are followed (Tyler, 2006a).

There are many situations where following rules is especially important, and one such situation is in a research setting. Ethical guidelines, procedures to reduce confounds and biases, and procedures for collecting and analyzing data must be followed to the letter by each individual in the lab in order to ensure data collected in an experiment is valid and reliable. However, not much research has been done on rule following in the lab; little research has been done in an attempt to translate the findings on obedience to a research setting, despite the benefits it could bring. Like Tyler (2006) and Sunshine and Tyler (2003) found, if an experimenter’s authority is seen as legitimate, they can rest assured knowing their rules will be followed even when they are not present, and they will not have to waste time and resources to ensure their rules are followed. Both of these would be very beneficial to a researcher, as time, availability and resources are limited.

So, how can one interpret the information that is available and translate it to a research setting? A study by Trinker, Cohn, Rebellon and Gundy (2012) found that stricter and less receptive parents, as well as parents that were too lenient, weren’t seen as legitimate, and thus their children were more likely to break rules. The somewhat strict but more receptive parents were seen as more legitimate, and so their children were less likely to disobey. To generalize these results, one could say that authority that is open, but firm is more effective than an authority that is too tough and distant or one that doesn’t take anything seriously. One must make rules designed to benefit not only themselves, but the individuals they hold authority over; and one must also be willing to explain why such rules are valuable and necessary, without deferring to, “because I said so.”

Another study by Fagan and Tyler (2005) found that for adolescents, their disobedient behavior was affected by how legitimate they felt the legal system was. Sunshine and Tyler (2003) and Tyler (2006 a and 2006 b) found that when people felt the legal system had legitimate authority, they reported accidents to police, had a favorable view of law enforcement, and were more obedient. These behaviors are favorable for a laboratory setting, since it’s desirable for students to follow rules, report accidents and view their mentor favorably. Thus the disciplinary system of the research setting must be seen as legitimate as well.

Other studies done on cheating behavior in students found that the amount of pressure a student feels to get a good grade effects cheating behavior (Barnett & Dalton, 1981) (Diekhoff et al., 1996) and that instructors who aren’t explicitly clear about what they perceive as academic dishonesty may cause instances of cheating to increase (May & Loyd, 1993) (Stern & Havlicek, 1986). While cheating is not totally synonymous to disobedience, the principle is similar. In a lab setting, fabrication of data for a better grade is also a concern; as well as students not being aware of what their mentor expects of them, or not knowing what to do because the information given to them was too difficult. A legitimate authority could minimize this by emphasizing that honesty is more important than having favorable data, and by clearly stating, more than once, what is and isn’t okay when handing data analysis. They would also have to be willing to work with students to ensure the information given to them is understandable. We predict that these results can be generalized in a laboratory setting, and a legitimate authority that is perceived as very strict and uncaring towards students will have more instances of rule breaking than a legitimate authority who is seen as somewhat strict but fair and open.

We cannot, however, assume that the results from these studies about parents, police and cheating can apply to a college research setting. In order to test our hypothesis, we conducted an experiment: 50 college students were placed into two groups under two authority figures, one “strict” and one “relaxed”. If results from previous studies can be generalized to a research setting, then the results should be that the stricter experimenter will be seen as less legitimate, and thus more rules will be broken; and that the less strict and more receptive experimenter will be seen as more legitimate, and thus, fewer rules will be broken. (Trinker, Cohn, Rebellon and Gundy, 2012).Method

Participants

Our participants were fifty college students from a local university, who agreed to participate to gain research credit. Among these 50 students, 20 identified as male, 19 identified as female, and 11 identified as non-binary/other. In terms of ethnicity, three were Caucasian, 10 were Asian, 11 were Hispanic, 13 were African American, and the remaining 13 identified as other.

Measures

To assess the students rule following behavior, (or obedience,) in each setting, several variables were accounted for. We measured the rule following behavior by frequency, (recording each individual time they broke the rules.) We defined this rule breaking behavior as each time a student fails to follow a rule; either by following a rule incorrectly, or by omitting a rule. There were 14 rules total, outlined in figure 1. A high frequency of breaking the rules, which would mean a higher numeric score, might indicate that a student broke more rules because they did not see the authority figure as legitimate. A low frequency of rule breaking, categorized by a low numeric number, suggests the participant saw the experimenter as legitimate, so they broke fewer rules. We also measured how the students felt about the experimenter to ensure the students interpreted the experimenter as we were trying to portray them. To account for other variables that could be responsible for any significant differences in our data, we also recorded the gender and ethnicity of each participant.

Materials

Participants were given a data collection sheet so they could record the behaviors of the fake subject they were told they’d be observing. The papers true purpose was for us to have a record of any rule breaking concerning data collection. We recorded a ‘fake subject’ sitting in a room with an unopened bag of skittles for twenty minutes, and then repeated the same video for each participant. While the participant was observing the video, a confederate used a one way chalkboard to observe the participant. The confederate was given the same data collection sheet so that they could record the participant’s behavior, as well as a sheet specifically for recording frequency of broken rules. The confederates were also given the same set of rules as the participants, so they could see each rule as it was given to the participants. The participants were given a bowl of skittles to stop them from being distracted by their own desires for the skittles. After the experiment, the participants were given an anonymous questionnaire in which they were asked to rate the experimenter. The purpose of the questionnaire was to check the students’ perception of the authority figure, to make sure we were portraying them properly.

Procedure

We randomly assigned each participant to the “strict” group or the “relaxed” group.Once the participants were placed, we sent them a forced choice questionnaire to obtain their gender and ethnicity, as well as their experience with, and knowledge of, academic research.Any student who indicated they did not have prior experience in research headed by university faculty, or that they did not have plans to participate in such research, were removed from the study. The participants failed if they were unable to answer more than half of the general research questions correctly, and were removed from the study. Any removed participant was replaced to keep the sample size at 50.

Before the night of the experiment, each participant was sent an email with the list of rules. All participants received the same document, with no differences. They were told to bring a copy to the study so the experimenter could go over it before the experiment began.Every participant brought a copy as asked.

The two experimenters were given instructions on how to act, and for certain key moments, each experimenter was given a script to follow. Their behaviors, word choices, and tone had to be controlled to represent each authority figure. The “strict” experimenter exhibited behaviors and attitudes of authority figures that are non-legitimate, and the “relaxed” experimenter exhibited the behaviors and attitudes of authorities that are seen as legitimate. To demonstrate this, below is a detailed account of how each experimenter treated their participants upon meeting them, and while explaining the experiment to them.The “relaxed” experimenter introduced themselves by name, offered to shake the hand of the participant, and described themselves as the head experimenter of the study. The “strict” experimenter, upon meeting the participant, simply said, “oh, you’re here, follow me.” Both experimenters led the participant to the room adjacent to where the experiment would take place, and asked them to sit down at a table. The “relaxed” experimenter first thanked them for participating and for bringing the document, then asked if they had any questions before going over the purpose of the study. The “strict” experimenter did not ask if they had any questions and did not thank them.

While explaining the purpose of the study, both experimenters followed the same script, but were told to maintain their “persona”. Both groups of participants were told they were not there as a subject of the research, but as an additional person to record data in order to minimize observer bias. The students were not “blind” to the fake experiment, and were told the point was to observe the subject’s behavior when left alone in a room with something desirable. Both experimenters gave the participants the same waiver to sign.

The “strict” experimenter was not given a script for this, and was told to maintain their “persona”. They covered each rule briefly, and skipped general rules that were not unique to the study, labeling them as “self explanatory”. They reacted negatively to any interruption, and made participants wait to ask questions. If the participant had questions afterwards, the experimenter answered them vaguely. If a participant asked them to explain why a rule existed, they answered with some variation of “because I said so,” such as “because it’s necessary for the experiment.”

He “relaxed” experimenter also had no script, and was instructed to maintain their “persona”. They were told to explain each rule fully and slowly, be open to on-topic questions and receptive to participants needs, stop at certain points to make sure the participant understood what they were saying, and explain the purpose behind each rule without being asked.

Once the experimenters finished explaining the rules, they both reminded the participant to follow the rules so that their data could be used in the study. The “strict” experimenter placed an emphasis on the need to get good data, while the “relaxed” experimenter emphasized honesty about errors, and told them they would not be penalized for mistakes. The participant was then left alone in the room, the confederate entered the room behind them to observe them, and the video was played. For twenty minutes exactly, the participant observed the fake subject and recorded their behavior, while the confederate collected data on any rule breaking that they observed.

Once the participant was finished, they were given a questionnaire to fill out which asked them questions about the strictness and effectiveness of the experimenter (figure 2).Once they finished, the experimenter explained the true purpose of the study. They signed a document agreeing not to talk about the study to anyone until its completion, and then they were sent home with their credit. Once all the participants had come in and the experiment was over, we collected the data the participant took and the data the observer took and began tallying the amount of broken rules for each participant, then used an ANOVA to interpret the data.

Results

We ran a one-way factorial ANOVA to analyze the data, and determine if any of the variables had a significant effect on the amount of rules broken. Our main hypothesis was that the authority figure being stricter would result in more broken rules, while a more relaxed one would have less broken rules. However, tests of between-subjects effects revealed that the data did not support this hypothesis; F(23, 26) = .400, P=.532. None of the other variables had significant effects on rules broken either: gender, F(23, 26) =2.208, p=.130; ethnicity, F(23, 26) = 1.002, p=.424; relaxed or strict * gender, F(23, 26) = .228, p=.798; relaxed or strict * ethnicity, F(23, 26) = 1.501, p=.231; gender* ethnicity, F(23, 26) = 1.324, p= 278; relaxed or strict* gender * ethnicity; F(23, 26) = .050, P=.951. Data from the questionnaire showed that 43 out of the 50 participants saw their experimenter “correctly” meaning they rated him or her how we would expect if the students perceived the authority figure as we meant to portray them.

Discussion

After examining previous research on obedience, such as the experiments conducted by Tyler (2006 b & 2006 a), Sunshine & Tyler (2003), and Fagan & Tyler (2005), we hypothesized that, if the results of such studies could generalize to college students in a research setting, then a more “strict” authority would be seen as less legitimate than a more “relaxed” authority. We tested this by recording the frequency of rules broken by participants during an experiment with two different groups. One group was under the guidance of the “strict” authority, and one group had a “relaxed” authority. We compared the results, along with gender and ethnicity of the participants, using statistical tests.

The analysis showed that the group the participant was in had no significant effect on rule breaking, and there was no significant effect between any of the variables. These results imply that neither the strictness of an authority, nor the identity of the participant, has any effect on how many rules a student may break in the lab. Meaning our hypothesis has been refuted, and the results of other studies do not generalize to students in a research setting.A possible explanation for these results is that we oversimplified the matter of obedience to address one specific factor. There are many reasons someone may choose to, or accidentally, disregard a rule. Matters that involve people’s personal choices are very complicated, and the focus of our experiment may have been to narrow. When looking at the individual pieces of data as a whole, the amount of rules broken by each participant seems completely random.This is most likely not because there is no pattern to such behavior, but that our experiment simply did not test any of the true causes behind such patterns.

A secondary explanation, which comes from the heart of our hypothesis, is that results from other studies simply don’t generalize to college students in a research setting. The studies we reviewed, such as Sunshine & Tyler (2003), Tyler (2006b), and Fagan & Tyler (2005), often used younger children, pre-teens, or adolescents in middle and high school to test their data. It’s possible that we have overestimated the similarities between students of different ages.

We picked participants who had prior research knowledge to account for a student’s lack of knowledge affecting how well they could follow rules. We tested two other variables, gender and ethnicity, in the case that one of them may affect the frequency of a participant’s rule breaking. We made it clear to the participants that they would be given research credit regardless of performance, so that students would not change their usual obedience behaviors to benefit themselves. However, in spite of these efforts, the data would suggest we did not go far enough to avoid various confounds. As mentioned previously, many factors go into an individual’s choice to follow or break rules, and we failed to account for enough of them. Our study also focused on a specific population, meaning our results may not generalize to other populations. We selected samples from our specific population of interest, college students engaged in academic research. We used a volunteer method of collecting participants, which could mean our results would only apply to students who would readily volunteer for research. Furthermore, these results could not be generalized to students with no prior knowledge of how research works, as we specifically removed those students from our participant pool.

Based on the results, it’s obvious the study suffered from poor content validity.Despite how well we set up the experiment, our experiment was not truly testing obedience behavior based on the legitimacy of the authority figure, but perhaps reading comprehension or one’s ability to understand rules on one’s own. While the two groups got different levels of support from the researcher, it made no difference in the data, implying that individual scores were caused by variables out of our control.

Had we done more preliminary testing, we may have been able to control for such causes. For example, it might have been helpful to test how much sleep the participant had the night before, as amount of sleep often has an impact on performance. We could have even designed this study differently. Perhaps as a within-subjects experiment, where all 50 participants would experience both authority figures by participating in two experiments instead of just one, would have been better. This would lessen the amount of variables needing control, because the individual differences of those in each group wouldn’t pose as much of a problem as in our original design.

Future studies interested in our hypothesis could see our results and infer that in obedience research, there are many variables that must be controlled and accounted for, and that a broader topic of interest may be more successful. It would be interesting, however, if a study chose to use our experiment design and test it in an environment that previous experiments have generalized to, such as a high school or middle school.

We hypothesized that the results of past research would generalize to college students working in the lab. If our hypothesis had been supported by data, research labs around the globe could have implemented less strict policies and receptive, effective methods of teaching students and enforcing necessary rules. While our data did not support our hypothesis, the idea that a strict and distant authority figure yields unfavorable results is a theory that has been well supported. Hopefully future studies will find answers to questions we were unable to answer, and the scientific community can continue to improve the world by improving its own research methods.