chapter 14 Flashcards
What are the 3 types of replication?
Direct replication ; conceptual replication ; replication plus extension
Give an example of direct replication.
In the case of toddlers effort when pressing buttons, researchers would replicate the same steps and procedures in order to compare results and see if the effect is the same in a new data collection.
Give an example of conceptual replication.
A conceptual replication explores the same research question but uses different procedures – there are important differences in methodological factors. For example, in the alcohol and aggression study, researchers can explore the same variables but operationalize them differently.
Give an example of replication-plus-extension.
In the study comparing handwritten notes and laptop notes, we could add different ways to take those notes, such as ewriter or no notes at all.
What are the two types of replication projects?
One study many labs ; many labs, many studies
Why might a study not be replicable?
It may not be replicable due to contextually sensitive effects (p.ex : the sample used is very specific such as SIPR students). There may also be problems with the way the original study was conducted.
What is a meta-analysis?
It is a statistical means of summarizing a number of studies using similar methods. We pool the effects to compare so we can make sense of the literature overall.
What are the strengths and limitations of a meta-analysis?
There is a file drawer problem and a publication bias.
What is a file drawer problem?
This occurs when we overestimate the true effect size cause null results and opposite results are rarely published
True or False : Journalists consider the importance of replicability.
False : Journalists tend to sensationalize one research finding.
What is p-hacking?
P-hacking occurs when researchers try many ways of analyzing their data, so the result is more likely to be a fluke rather than a true, replicable pattern. P-hacking is often used in order to ensure publication of data (avoiding null effects).
How can we avoid P-hacking?
P-hacking can be avoided with open data, in which full data sets are provided. Others can then re-run and confirm the statistical analyses.
What is harking?
Harking is when the study reveals and unexpected result, but the researcher writes about the study as if the result had been predicted all along (writing the hypothesis after doing the study).
How can we avoid harking?
Harking can avoided with preregistration, in which researchers publish the hypothesis and study design before data collection and analysis begin.
How can small sample sizes be problematic?
In a small sample, a few chance values can influence the data set, so the study’s estimate is imprecise and less replicable.
