Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
16 Cards in this Set
- Front
- Back
|
What's the biggest issue when trying to conduct experiments in real life? |
The inability to separate participants into randomly assigned groups. |
|
|
The types of statistical tests put into place in quasi and experimental designs tend to be the same/different |
The same. |
|
|
Define a one group post test only. What type of experiment is this? What does it prove? (If anything) |
A one group post test only is taking one group, giving them the independent variable treatment, and then measuring the after effects. This is a quasi-experimental procedure, and it can only (barely) point towards correlation, because we don't have a control group, and we never measured the condition before the test. |
|
|
Quasi experimental groups will take measurements how often before and after a intervention (treatment) |
A lot of times! |
|
|
What's a quasi-experimental study involving naturally-occuring treatments? How can a researcher prepare the pre-tests for this? |
It's when the researcher wants to understand the effects of a naturally occurring disaster or something. It's impossible to specifically prepare pre-test data!! Researchers usually rely on archives that already exist and refer back to them when a natural disaster happens! |
|
|
What's a group differences study? What type of info do we glean from this. |
A group differences study is a form of quasi experiment, where the experimenter gets two naturally occurring groups and DOES NOT administer a treatment or intervention. Eg. Male vs female reaction time. This can prove correlations but never causation (it's a quasi!)
|
|
|
What's a median split? What does it do? What are some problems with it and how do researchers avoid it? |
A median split is used when a researcher needs to made 2 naturally occurring groups based on characteristics that are less obvious (extraversion, IQ). A median split is when all participants are tested for the quality, and then the top 50% go into one group, and the bottom 50 go into the other. This makes 2 groups that aren't very different (49 and 51 are on different ends) so sometimes researchers will use the top and bottom 25% instead, to make the groups more different. |
|
|
What's a cross sectional design, and give an example |
A cross sectional design is taking people of different quasi-independent variables (age?) and measuring them at the same time.
|
|
|
What's a cohort? |
A grouping of people born around the same time. |
|
|
Longitudinal designs are... |
Kinda the opposite of cross sectional. You take a group at time 1 and then measure them at many different time intervals. |
|
|
What's the cross generation effect and what does it mean? |
The cross generation effect means that it's hard to generalize results between generations, giving longitudinal research designs lower external validity |
|
|
What's a secular trend? |
The idea that conditions and quality of the world change over time, which effects longitudinal research designs and really threatens internal validity of these types of study. |
|
|
Attrition is a threat to internal validity of a longitudinal study. Explain what attrition is, when it's especially bad (and not as bad), and why it's specifically annoying in longitudinal studies |
Attrition is the drop-out of participants in a study. It's bad when the drop outs are for specific reasons, and not so bad when it's purely random. It's really annoying in longitudinal studies because it takes SO long to collect data. |
|
|
What's the temporal precedence effect? |
Temporal precedence means that things that imply causation have to happen in the order that the cause MUST occur before the effect. |
|
|
What are Cross-Lagged panel designs? |
A way of showing a variable on top at time one and time two, and the second variable below it laid out the same way. You draw lines between each and it shows the way the variables are interconnected. |
|
|
Name the three types of correlations in the cross-lagged panel design. |
Synchronous (V1 at T1, V2 at T1) Autocorrelations (V1 at T1, V1 at T2) Cross lagged correlations (V1 at T1, V2 at T2) |
