Unit 2 Flashcards
(45 cards)
the goal of a study
to be unbiased and provide good results
EPSEM
equal probability of selection method
household bias
when a respondent answers for a group,
over represented individual
stratified random sampling
population is divided into groups called the strata
how?
divide population into strata
randomly sample each group in proportion to size (household bias)
why?
can provide more precise results
groups are different but are similar within
types of bias
sampling bias
non response bias
household bias
response bias
5 types of causal relationships:
cause and effect-one changes, effects other
common cause factor-external factor effects both
reverse cause and effect- opposite independent and dependent variables
accidental relationships- not effecting each other only coincidence
presumed relationship- correlation seems to exist but difficult to identify a common factor
what are the two types of data sources?
primary- collected by person doing analysis
and secondary- collected by an outside source
index
number that provides a measure of scale to compare data
ex. Rictor scale measures earthquakes
in a survey you should…
avoid the truth
uphold a high standard for your work
what is bias?
an unintended influence on a. data gathering method
what is a thesis?
the focus of your project that declares what you believe and are proving
other sampling methods
destructive sampling-samples destroyed in the process
voluntary response sample-people invited to participate, often skewed results because only some respond, not random
convenience samples- not random, pick random people off the street
what is sampling bias?
when the chosen sample does not accurately represent the population
inference
a conclusion on the population based on sample data
quantitative data
variables measured numerically
what should you not do in a thesis?
avoid vague language
not first person
in Excel what does the dollar sign mean?
absolute or fixed
complete the same operation by the same number each time
if statement in Excel completes the action:
if one is greater than a certain value a ___ will be shown if it is less a ____ is shown
Hawthorne effect
people behave differently when being watched
correlation
the apparent relationship between two variables
positive/negative, strong/weak
placebo effect
an inactive drug is administered to a subject who exhibits a positive reaction to the drug
what makes a good thesis?
be specific and focused,
should propose an arguable point
simple random sampling
all selections and combinations are equally likely
deviations due to chance
how?
generate random numbers
assign the numbers to all elements in population
pull numbers at random and choose those elements
why?
easy to understand
theoretical fairness
census
a collection of data taken from the entire population