Summary
Summary MCRS exam 2
- Course
- Institution
Here's a summary of all the lectures, micro lectures, readings and tutorials of MCRS for exam 2
[Show more]
Seller
Follow
celiataboadagarzn
Content preview
MCRS exam 2week 8
1.1 Internal/external validity threats
Ecological isomorphism: The extent to which experimental conditions are similar to the outside world the experiment
seeks to replicate. Therefore, the experiment may have questionable validity.
Internal --> Threatened if there's a plausible alternative explanation for the study results
Types of threats due to participants:
● Maturation
- Alternative explanation formed by natural change
- i.e. loneliness causes depression so a cat is given to reduce it, if depression is reduced, it could be
because of natural change instead of the cat, people sometimes change
- Solution: introduce a control group
● Selection
- Systematic differences in participants characteristics
- i.e. People in the control group can't move and their depression levels were higher than the experimental.
Low levels in the experimental may be because they are more fit instead of the cat.
- Solution: random assignment
● Selection by maturation
- Groups systematically differ in their rate of maturation
- i.e. experimental group volunteers and control group conservative people. I experimental group has less
depression after the cat is given, it could be because of their "lifestyle"
- Solution: randomized assignment groups
Types of threats due to instruments:
● Low construct validity
- Low if our instruments contain a systematic bias or measure another construct or property entirely
- Construct validity is a prerequisite for internal validity
- i.e. H1: by giving them the cat, it didn't affect loneliness but instead it affected their social status making
it higher. The manipulation aimed at lowering loneliness in fact changed a different property: social
status. Moreover, the questionnaire to measure depression instead measured social acceptancy so we
cannot draw any conclusions about the relationship loneliness and depression.
- solution: valid instruments and valid manipulation methods used consistently
● Instrumentation
- When an instrument is changed during the course of the study
- i.e. self report questionnaire to measure depression at the start of the study but an open interview at the
end of the study. Then, any different scores might be explained by the use of other instruments
- Solution: valid instruments and valid manipulation methods used consistently
● Testing/sensitization
- Administrating a test can affect people's behavior and provide an alternative explanation for results
- i.e. taking the depression pretest at the start of the study might alert people to their feeling of depression,
this might cause them to be more proactive about improving their emotional state for example by being
more social.
, - Solution: introducing a control group. Both groups will be similarly affected by the testing effect, their
depression scores will both go down but hopefully more so in the cat companionship group.
- Adding a control group is not always enough though, in some cases, there's a risk that the pretest
sensitizes people in the control group differently than people in the experimental group.
- i.e. the pretest in combination with getting a cat, could alert people in the experimental cat group of the
purpose of the study, they might report lower depression on the post-test not because they are less
depressed but to ensure that the study is successful so they can keep the cat (alternative explanation)
- Solution: add an experimental and control group that weren't given a pretest
- Basically use a special design that includes groups that are exposed to a pretest and groups that aren't
Types of threats due to artificiality:
● Experimenter expectancy
- If the researchers expectations have a biasing effect. Unconscious change in the researchers behavior
caused by expectations about the outcome that influences a participants responses
- i.e. Rosenthal 1950s. Psychology students were led to believe that they were taking part in a course on
practical research skills in handling animals and duplicating experimental results. Students handled one
of two batches of rats that were compared on their performance in navigating a maze. 1 batch was bred
for their excellent special ability, the other one for the exact opposite purpose. There was a difference in
performance, the maze bright rats outperformed the maze dull rats. In reality there was no difference in
them, they were both from the same breeding line, just students treated them differently.
- Solution: experimental blind design to minimize subjective bias
● Demand characteristics
- Change in participant behavior due to the expectations about the study.
- i.e. people are aware that they are in an experimental group and are undergoing a treatment, specially if
the treatment aims to help them. Participants might be grateful to be in this group or hopeful that the
treatment will work. This might lead them to be more positive about the treatment.
- Solution: double-blind research design: both participant and researcher are unaware of which participants
are in each group
- Cover story: plausible explanation of the purpose of the study. It should provide participants with cues
that are unlikely to bias their behavior
External/generalizability--> If the hypothesized relation holds for other persons, settings and times
History threat
- The observed effect doesn't generalize to other time periods
- i.e. compliance study performed in the 1950s in the US, results showed that participants were willing to
comply with highly unethical directions provided by an authoritarian experimenter. This results will be
less extreme if we conducted the experiment nowadays because for example, people are more educated
and less sensitive to authorities.
- Solution: replicating the study in a different time or by repeating the study in different settings.
Setting threat
- The observed effect only hold in a specific setting, the findings did not generalize to other
environments or situations
- i.e. investigate the relation between violent imagery and aggression and find that children who watch
violent videos are more aggressive in the playground afterwards. A setting threat will happen if the
children only behave badly in the playground but not at home.
, Setting threat artificiality
● Pretesting
- The observed effect is found only when a pretest is performed. Closely related to the internal
validity threat of testing.
- i.e. investigating a new therapy for treating depression and use a pretest. The depression pretest
makes participants realize how serious their problem is and thereby makes them more receptive
to treatment. The treatment is effective, but only if the receptiveness is increased by the pretest
first
- Internal validity is threatened because receptiveness is missing from our hypothesis
- External validity is threatened because the hypothesis will only apply to situations where a
pretest is part of the setting
- Solution: repeating the study in a more natural environment
● Reactivity
- When participants or experimenter react to the fact that they are participating in a research study.
- Includes participant and experimenter expectancy and altered participant behavior (i.e. due to
nervousness). This can cause the hypothesized relation to occur only in a research setting and
not in a natural setting
- i.e. Investigating a new method for teaching high school math. The researcher is present during
the lessons and measures math performance in class. What if students worked harder because
they know that they're being studied? And this makes the new method more effective.
- In a natural setting without the researcher present students might put less effort, reducing the
effectiveness of the new method.
- Solution: repeating the study in a more natural environment
Selection threat
- When the hypothesized relation only hold for a specific subset of people or if the results in our study are
biased due to over or under representation of a certain subset
- i.e. in our study of a new depression therapy we recruited participants who actively volunteered and we
find that the therapy method is effective. The volunteers will be more proactive in solving the problem
that an average person is. The method will be more effective in volunteers than other people.
- i.e. asking people at a university for women's right to vote. Sampling is so selective that it's unlikely that
we can generalize the results to the generalized publics opinion.
- Solution: repeating a study with different groups of subjects or conducting a random sampling of the
research sample , which is called probability sampling
1.2 True experiments
If…then… best hypothesis for an experiment
- Experimental research designs maximize internal validity, they're referred to as true experiments or randomized
control trials/RCTs. Best defense against alternative explanations for causal claims
Manipulation
- If you want to show a causal relation, the strongest possible empirical demonstration is when the cause is
under your control, this way you can show that the cause precedes the effect, eliminating an ambiguous
temporal precedence
Comparison
- Causality is even more plausible if you can compare to a situation where the cause is absent, showing that
the effect does not occur when the cause is absent, this also eliminates the threat of maturation. Ensures
that the effect did not occur naturally
, Random assignment
- Randomly assign children to the experimental and control condition. It ensures it is equally distributed
and that there is no systematic difference between the groups other than the IV.
- Randomization can fail sometimes, the only way to make sure it works is by replicating the study.
1.3 Within subjects design
Between subjects design: when we investigate an IV that can be manipulated the approach is to expose different groups
of participants to the different levels of IV. The IV is now called a between (subjects) factor
Within subjects design/repeated measures design: let all participants experience all levels of IV. The IV is now called a
within (subjects) factor.
i.e. investigate the effectiveness of an experimental drug in reducing migraine attacks at different dosages
- Standard approach would be to assign randomly each patient with a low medium or high dose for one
week (between)
- We could also choose to let each participant experience all 3 dosages (within)
Within factors can be combined with other between or within factors in a factorial design.
- i.e. In addition to the factor dosage, we can also investigate the factor gender. As you can't be exposed to
both male and female, it's a between factor.
- i.e. a group of men is exposed to all three dosages and the same goes for women--> between factor
gender combined with a within factor dosage
Within factors:
Design is more practical as you don't need that many people in your study
Repeated measures design
- DV measured repeatedly.
- Same participants are measured on the DV after a short period of time and after being exposed to each
level of the IV, otherwise we wouldn't know what the effect each level or condition is.
- At least one within factor and possibly (multiple) between factors
Longitudinal design
- DV measured repeatedly.
- Studies that measure the same variables repeatedly but over a long period of time, mostly correlational
studies where no IV are manipulated.
- It includes experimental or quasi-experimental studies
1.4 Manipulation
- Usually refers to control over the IV.
- The value or level of the IV is determined and manipulated by the researcher.
- It helps to have control over external variables by keeping variables of disinterest constant we can rule out any
alternative explanations
Manipulation of the IV:
- i.e. H1: violent imagery is a direct cause of aggression. To test, we can manipulate IV by letting
participants play a violent videogame for either 2h, 4h, or none. We've created 3 levels of IV.
- If the IV is absent (none hours) is called the control condition/group
- If the IV is fully controlled by the researcher is called experimental variable
- Intrinsic property of a participant that cannot be controlled (age, gender) is called Individual
differences variables
The benefits of buying summaries with Stuvia:
Guaranteed quality through customer reviews
Stuvia customers have reviewed more than 700,000 summaries. This how you know that you are buying the best documents.
Quick and easy check-out
You can quickly pay through credit card or Stuvia-credit for the summaries. There is no membership needed.
Focus on what matters
Your fellow students write the study notes themselves, which is why the documents are always reliable and up-to-date. This ensures you quickly get to the core!
Frequently asked questions
What do I get when I buy this document?
You get a PDF, available immediately after your purchase. The purchased document is accessible anytime, anywhere and indefinitely through your profile.
Satisfaction guarantee: how does it work?
Our satisfaction guarantee ensures that you always find a study document that suits you well. You fill out a form, and our customer service team takes care of the rest.
Who am I buying these notes from?
Stuvia is a marketplace, so you are not buying this document from us, but from seller celiataboadagarzn. Stuvia facilitates payment to the seller.
Will I be stuck with a subscription?
No, you only buy these notes for $5.43. You're not tied to anything after your purchase.
Can Stuvia be trusted?
4.6 stars on Google & Trustpilot (+1000 reviews)
62890 documents were sold in the last 30 days
Founded in 2010, the go-to place to buy study notes for 14 years now