Validity and Reliability of Milgram PDF

Preview

Summary

This is a checklist for everything that may come up in the exam from sections A, B and C for the Component 2 exam of Psychology. ...

Description

Say What Validity and Reliability Milgram had in his Experiment and Explain Where it was Shown

Reliability Issues

Internal Reliability: This is where the results across the test are consistent. An example of this in Milgram’s study was that 26 people out of the 40 completed the experiment. Another example was that most of the participants needed prods to continue. This was used because it was able to prove his hypothesis and show that on general, all American men will get the same results so not each individual had different results. External Reliability: This is where you get the same results on different days in different labs with different researchers. Milgram had this because he did the experiment at Yale university with science-looking experimenters and he also did it in an old, run-down mill with an experimenter who didn’t wear a lab coat. However, he did not find the same results. This was because participants were responding to authority so didn’t respond as much when there wasn’t a clear authority figure, Inter-Rater Reliability: This is where the researcher’s behaviour and the experiment should stay the same. An example of this in Milgram’s experiments was the fact that when it was held at Yale, the experimenter dressed and acted the same, they used the same prods, it was set up in the same way, the voltages were the same and the order of the method was the same. This means that every participant experienced the same test so there will be no extraneous variables. Inter-Observer Reliability: This is where the observers have to agree what they see. This was not included in Milgram’s experiment because you only had one observer and it was the same observer in each experiment so he had no one else watching so he didn’t have to compare answers. Because the experiment had both qualitative and quantitative data, a second experimenter was not needed because the data collected was numbers or reactions which could only be interpreted in one way. Test-Retest Reliability: This is where participants take the same test on different occasions. Milgram doesn’t have this because he only tested each participant once so you cannot measure with a retest if it was a different participant because it is likely that they would have very little similarities so their results couldn’t be compared. He did however, do the experiment on different occasions but kept them the same. Split-Half Reliability: This is where you compare a participants’ performance on two halves of a test. You cannot do this in Milgram’s experiment because they only do the experiment once and you could not

do it twice because they find out about the deception. You could, however, look at their qualitative data before 300V and after 300V and see what difference it made to their reactions when the learner didn’t respond to the question.

Validity Issues

Internal Validity: This refers to factors such as extraneous variables or bias that could lead to a change in attitude. Bias would not be likely to occur because the teacher had no idea what was going on because they thought they were testing the learner. The only possible extraneous variables that could change anything could be hunger or tiredness or could be what mood they are in to whether they spare their learner or if they are tired so aren’t as aware what is going on. Face Validity: This refers to if the test measures what it is supposed to measure. This is not in Milgram because the participants were deceived so they did not know what was going on else it would have changed their results. They thought the learner was strapped to the electrocution chair but they were not. They also thought that it was the learner that was being measured but it was actually them being measured. Construct Validity: This is the extent to which a test relates to an underlying theoretic concept. Milgram’s experiment related to two underlying theoretical concepts. The first one is the nature vs nurture argument and this proves that this theory is wrong. The other one it relates to is the ‘Germans are different’ hypothesis and this experiment proves that Germans are not different because most people would obey an authority figure if told to. Concurrent Validity: This means that test scores are checked against other similar, valid tests. Because this was the first test of its kind, it cannot be directly linked to any other tests. However, it could be linked to other tests such as Bandura et al. because this experiment also tests responses and obedience to an authority figure and the participants are asked to do something that they know are wrong but still do it because they are told to do so by a higher figure. Predictive Validity: This means the extent to which a test predicts an outcome in the future. You could say that Milgram’s experiment did this because he could see through his research and findings using different Independent Variables, that people will respond to a high authority figure. This could then predict that if a strong leader proposed something in the

future, even if it was wrong, people would still do it because of their authority. External Validity: This refers to the extent to which the results can be generalised to other places. This means that would you see this experiment in other situations. In this case, you would not because you are unlikely to find yourself at a prestigious university in America testing how a learner would react to a strong authority figure when they were being told to do something that was unethical and potentially harmful.

Ecological Validity: This refers to if this is a natural occurring experiment. Milgram’s experiment has a very low eternal validity because it is not like real life. The test is not something that you would see reoccurring and neither are the method measures. On the other hand, you would see this type of strong authority in everyday life so this factor of having to follow an authority figure could show an external validity.

Population Validity: This means if we can generalise our findings from our research to other groups of people. Because Milgram had a very specific person that he wanted, he only tested males between the ages of 20-50 so you could only generalise people of this gender and age. In separate experiments he also tested women so you could generalise it but could not really compare it because it was a specific type of person....