Statistics 1-4 Flashcards
What are three reasons to study statistics?
being an information consumer
understanding/making decisions
evaluating decisions that affect your life
Descriptive statistics
Methods used for Describing a data set
Inferential Statistics
The generalization from sample to population
variability
There is always natural variability
Variable
The things changing from individual to individual
univariate versus Bivariate data
Depends on the number of variables, univariate= one variable
categorical data
QUALITATIVE: relating to different categories
-color
-shape
numerical data
QAUNTATITIVE: relating to different numbers
discrete- 1,2,3
continuous- 1-3
Steps for planning/conducting a study
- understanding the nature of the problem
- deciding what to measure and how
- collecting data
- summarizing data
- analyzing data
- interpreting results
Confounding Variable
A variable the may be the cause of differences in the response variable because of its relationship to the population/sample
Selection Bias
when the conductor of the experiment chooses the subjects
wording bias
cherry picking your words
nonresponse bias
out of a randomly selected group people choose to answer
response bias
when a subject lies or doesn’t know the answer
measurement bias
multiple different people sampling, and using different methods or materials
volunteer response bias
When EVERYONE (or a whole population) is given the option to do an experiment
volunteer response bias versus non response bias
volunteer response bias applies to a whole population while nonresponse bias applies to just the sample
under-coverage
The effects of bias
house hold bias
___
Sampling
The methods used to select a sample from the population
SRS (simple random sample)
completely random sampling
-random number table
Stratified random sampling
When the population is split into categories and individuals are randomly selected from each group
Convenience sampling
When the sample is chosen just by what is easiest
Volunteer sampling
When people choose to take part in the study
Quota Sampling
Similar to stratified random sampling
-split into groups and percentage of individuals is taken from each group
Multistage cluster sampling
Breaking people into random groups and randomly selecting a few groups. Then randomly selecting smaller groups, and then individuals.
(has multiple stages)
What are the key concepts of experimental design?
randomization
blocking
direct control
replication
placebo versus nocebo affect
the placebo effect is shown when someone gets better just because they think they’ll get better, the nocebo effect is when someone feels bad because they think they will feel bad.
