Drawing Conclusions from Studies Flashcards
What is a null hypothesis?
A hypothesis suggesting there is no association to be observed, all visible correlations are caused by error (explanatory variable has no impact)
What is the structure of a hypothesis?
If-then statement; If (observation unit) is changed by (statement about explanatory variable) then I expect that (statement about change in the response variable)
What is an alternative hypothesis?
A hypothesis that reflects the expected change in the response variable due to the explanatory variable
What is an experiment?
A study that imposes treatment or changes the state of the observational units
What is an observational study?
A study using previous data or observing and recording information without imposing a treatment or changing the states of the observational units
What is a population?
All observational units in a certain category
What is a sample?
Some of the observational units in a population
What is a parameter?
Measures of a population
What is a statistic?
Measures of a sample
What is a representative sample?
A sample with similar characteristics to its original population
What is the sample size?
The number of observational units sampled
What is sampling frame?
A list of all observational units in a population
What causes a sample to be random and fit Simple Random Sampling (SRS)?
All observational units in a population have equal chance of being selected
What is convenience sampling?
Data is collected from observational units the researcher has easy access to
What is voluntary response sampling?
Submitting data for a study is optional (e.g., a survey that is not required)
What is undercoverage bias?
Parts of a population are less sampled that others in a data collection and underrepresented
What is non-response bias?
When those selected for a data collection refuse to respond and are underrepresented
What is response bias?
The behavior of the observational unit or researcher causes a bias in the data (e.g., people are already biased for or against something when interviewed and those are the only people interviewed)
What is wording bias?
A question written poorly enough to confuse the audience in their response
What is a lurking variable?
A variable that is not the explanatory variable but is related to the response variable and might be the cause of an association between the explanatory and response variables
What is a confounding variable?
A variable that explains the explanatory variable and the response variable and explains the association between them (the explanatory variable does not cause the response variable, the confounding variable does)