PSYC 3480 Lecture Notes - Lecture 2: Random Assignment, Internal Validity, Shared Belief
10 Jan 2016
School
Department
Course
Professor
PSYCH 3480 – LECTURE 2 2015-09-22 5:31:00 PM
10:15-11:15 am before class Deborah TA
additional info – from textbook
handouts
RESEARCH ISSUES
CLASS OUTLINE:
MOST OF THE INFO → is not in the textbook
Stereotypes – killing us softly III
theoretical perspectives
RESEARCH: (slide 4)
why is it important to talk about research – from the context of our class
reading – in the handouts + textbook - # of studies
How you can evaluate the research
Can know if you can dependent on the research
Good vs bad research:
Good Research: (What is good research?)
- Research that is ethical
- reliable/repeated
- good research – that is RELIABLE (replicable), FREE OF BIASES (VALID)
-
- Results should be very similar
- valid – what does it mean to be said that the research is valid?
• – is it measuring what its suppose to measure (internal validity).
• Generalizable – that is applicable to everyone
BAD RESEARCH:
- research that is invalid and not reliable
example of bad research: (hand out) – when good research goes bad
find more resources at oneclass.com
find more resources at oneclass.com
→ post menstrual hormones – hormones stop producing + perioids stop.
Some women not all have severe side effects – drug companies –
medications for this. To make it sellable – they had to test it
had a large sample test – hormonal replacement → half the women did not.
Women who took the replacement hormonal therapy – they not only did
they have their hot flashes go and but they also had better hearts. Drug
companies → put out all this info. THE SIDE EFFECT – was that it had
chances of breast cancer within it.
After 10 years – they tested the drug again → the women who took the
hormonal therapy replacement – they had worse HRs. Do not give to
everyone and only give to people for loss of hormones and for a short period
of time.
What happened? how could the 1st study say hrs are getting better but 2nd
study say they are getting worse?
→ took hormone replacement therapy – caused breast cancer
→ research – looking at the possibility of causing this. Sample is a part of it
and methodoly that had shifted. The randomization
1st study: women who took the hormone replacement therapy – had better
hearts and were health y and young
2nd study – they used randomization – they used random assignment –
everyone has a equal chance for hormone replacement therapy. They were
not biased samples. The study showed a true picture and hormone
replacement therapy did not help the HR
THUS THERE WAS BIASED SAMPLE in the 1st study.
This drug cannot be used for general health and HR
the problem of the research → with 1st study – was there no – random
assignment that was present
SLIDE 5: RESEARCH AND BIASES IN GENDER RELATED RESEARCH
what is asked? - asking relevant questions like related to men and women
how is it asked?
Looking at - domestic violence – divorce and why people were divorcing?
Had a pilot study – one of the questions that were asked?
one of the questions were asked was – are you afraid of your ex-partner?
RESULTS:
find more resources at oneclass.com
find more resources at oneclass.com
of the ones that did have fear – a lot of women → using the likret scale.
but something that was confusing was – that men were also scared and
fearful.
Men - had dropped down to along nothing - after a year.
In the questionnaire → why? As seeing the reasons – about the results.
the change – was HOW THE QUESTION WAS ASKED – WHAT DIDN’T WE
DO? – did not clarify why they were feared. What was meant by fear was not
defined. In the study – the fear was more physical and domestic violence
related.
men – thought they might loose their children, embarrassment at work,
having to financially support.
→ thus there must be an OPERATIONAL DEFINITION - to make people
understand what context it is being talked about
HOW THE RESEARCH IS CONDUCTED:
1. must do an experimenter expectations
- the experimenter should not let the participant know the results they
expect from them
- being aware of having the experimenter expections
2. research participants:
- who?
• Example of a study: A family went to camping. Thalidomide – for
nausea/early morning given to women who are pregnant
• women had babies with deformities
• problem with the study: thalidomide → were only tested on men. It
was not tested on women
• who were the things tests on? That was what the problem was –
thalidomide.
• Who – just tested on men but given to women
WHY?:
• reason why they didn’t test? – probably b/c of money (money
offered – ordinarily take the med would take it).
• mandatory
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
10:15-11:15 am before class deborah ta additional info from textbook handouts. Most of the info is not in the textbook. Stereotypes killing us softly iii theoretical perspectives. Research: (slide 4) why is it important to talk about research from the context of our class reading in the handouts + textbook - # of studies. Can know if you can dependent on the research. Good research that is reliable (replicable), free of biases (valid) Valid what does it mean to be said that the research is valid: is it measuring what its suppose to measure (internal validity), generalizable that is applicable to everyone. Research that is invalid and not reliable example of bad research: (hand out) when good research goes bad. Post menstrual hormones hormones stop producing + perioids stop. Some women not all have severe side effects drug companies medications for this.
