Rebecca Johnson PSYC37- Chapter 7 notes October 1, 2013
The Examiner and the Subject
Both the behaviour of the subject and the examiner can influence the test scores.
Eg. Children in grades 1-7 were given the WISC test, half the students were under an
enhanced rapport setting and the other was in a normal setting. In the controlled setting
the test examiners talked to the students more and were more warming. In the results it
showed that it did not effect children in grades 1-3 but IQ of 5 -9rh graders were higher
for those students who were in the controlled setting.
In another test where there were 2 setting –approving comments- and disapproving
comments. The setting where the students received disapproving comments scored lower.
Being familiar with the test taker can also alter results. About 4 points higher on IQ.
Substantial interviewer effects: a respondent might give a response that they think the
interviewer wants. This is known as response acquiescence, and causes systematic bias.
Examiners should know that their rapport with the test takers can influence the results.
The Race of the Tester
Some people feel that the race of the tester effects the results
Some African American parents feel like their children score lower on IQ tests when
tested by a white examiner.
Sattler concluded that there is not a lot of evidence that supports this information
The most common findings is that the race of testers does not affect the performance of
both AA children and white children
Effects have only been seen in 4 out of 29 studies
The reason why only a few studies show effects is because when administrating an IQ
test the rules are so strict and specific there the race of someone does not really matter. In
reality if an examiner follows the proper steps they should all act the same.
An explanation to why AA children score lower could be because they have trouble with
reading in one study where the test was listened through audio tape, the AA children
scored significantly higher on the test.
Language of the Test Taker
The amount of linguistic demand can put non English speakers at a disadvantage
Translating a test can be difficult because it can cause bias
It cannot be assumed that the translation is completely valid and reliable in comparison to
If someone speaks multiple languages they should be given the test in the language that
test taker feels it is their best
Having a translator can cause bias and should only be used with great caution Training of Test Administrators
Different assessments require different levels of training from the tester
Some require a professional degree others do not
There are no standardized protocols for training people to administer complicated tests
like the Weschsler Adult Intelligence Scale- Revised (WAIS-R)
In a study with 22 graduate students, many errors were found in scoring the tests; the
error rate went down after about 10 administrations. This suggests that students need to
practice administering tests about 10 times to begin gaining competence.
Research shows that sometimes data can be affected by what the experimenter expects to
Robert Rosenthal- from Harvard University , conducted many experiments on
Expectance effects, also known as Rosenthal effects
Typically what would happen is that a group of students would gather to collect data, half
were told that the average response would fail and the other half were told that the
average would fall towards the success side.
The results show that the subjects would actually provide data that confirmed the
This has also been seen in experiments with rats, either being maze dull or maze bright.
The rats were all from the same litter they were just labeled one or the other.
Many authors have challenged Rosenthals experiments stating that they are bases on
unsound statistical procedures or faulty design.
A review of many studies suggests that expectancy effects is present but not in all
Expectancies influence our judgement in many ways eg. When reviewers are looking at
candidates for grants. Reviewer’s expectancies about the investigators do influence their
2 aspects of the expectancy effect relate to the use of standardize tests
1) The expectancy effects observed in Rosenthals experiments , were
obtained when all of the experiments followed a standardized script
2) The expectancy effect has a small and subtle effect on scores and can
occur in some situation but not in others.
The expectancy effect can impact intelligence in many ways like scoring
Examiners need to be aware that their relationship to interaction with the examinees can
affect their objectivity when they score certain types of tests.
It is very important to do as much as you can to eliminate the possibility.
Effects of Reinforcing Responses Reinforcement affects behaviour, therefor administers should test under s controlled
Inconsistent use of feedback can damage the reliability and validity of test scores
Rewards can significantly affect test performance eg. Incentives improve test
Many studies have shown that children work hard to obtain praises or rewards
The type of praises is important as well
the results of the studies are complicated
Culturally relevant feedback also does play a role. In AA children they showed higher
results when given culturally relevant feedback versus other kinds of feedback like
Some of the most potent effects of research studies are seen in attitudinal studies.
The answer is not right or wrong but how much someone feels about something.