Psychology Lecture 7
Slide 16: Reliability- you look for precision for an instrument to be reliable. (precision- get consistent
numbers when you repeat the same action)
Validity- a valid measure of what it says it measures, but it’s not reliable. Ex. Measuring height using
elastic band is valid, but not reliable, because you can stretch it.
The example of Lecture 6 (IQ b/w white and black kids): IQ tests seemed reliable, but after that incident,
it was not. You also test the person’s reading and comprehension when you use words, but the result
could vary among the person and their background. You measure more than just intelligence.
Ex. You wanna study a baby- at what point would they see colour? Use preferential looking task: two
monitors and a camera in front of the baby- one coloured, one B/W, and they change colour and black
and white screens between those monitors. The baby should look at the colour picture since it’s more
interesting. However, babies do not follow those directions, they do whatever they want- they’re still
looking, but they are disguised under their drooling and bobbing etc. So how can you make this reliable?
Independent scores: research students watch these experiments and score where they think the
babies looked and compare with each other’s record. If these are the same, then there’s more
confidence. Inter-rater Reliability
Are there dependent/independent variables? –independent is difficult to categorize (age), since
you pick babies and just use the age that they are in.
Slide 17: For imprecise experiments, judges come and decide and agree what is happening.
Slide 19: An experimenter was an attractive looking man, many women looked at him, and in
relationships would be more short-time.
Control experiment: participants use computers, and the experimenters are present as well to see their
self-consciousness. The experimenter chooses where the participants go once they enter- the attractive
women were put into the room where the experimenter was there as well (he was attractive as well).
Experimentally that was bad; the goal was to see the difference of being watched and not watched. But
the watched groups were predominantly women, and most likely were talked to. This would mean that
other groups could have had different results. he created a confound (anytime when groups differ
other than the ones