exam 1 Flashcards
accuracy
getting close to right value
precision
getting right value multiple times
what is the goal of sampling
minimize sampling error and minimize bias
sampling error versus bias
error- precision
bias- accuracy
random sample
each member of population has an equal and independent chance of being selected
if random sampling minimizes bias it therefore,
increases accuracy
large samples increase what
precision but do not affect accuracy
what are the 2 ways to achieve poor samples
sample of convenience and volunteer sample
in terms of categorical variables what are the 2 ways they can be subdivided into
nominal and ordinal
nominal- no inherent order
ordinal- order
in numerical values what can they be subdivided into
continuous- weight, temperature- can take on any real. number value
discrete- no halves, numbers on a dice
what is the explanatory variable
predicts/affects the value of the other variable - independent variable
what is the response variable
potentially affected by explanatory variable - dependent
experimental study versus observational studies
experimental- can identify relationships, cause and affect
observational- only associations. no manipulation
when using numerical data what kind of graphs
frequency table and histogram
when using categorical data what kind of graphs
frequency table and bar graph
if showing association between 2 categorical values what kind of graphs
grouped bar graph, mosaic plot
if showing association between 2 numerical values what kind of graphs
scatter plot, line plot, maps
association between a categorical and a numerical variable graphs
violin plot, box plot, strip chart, multiple histograms
how to show trend in time
line graphs
how to find the sample mean
sum of Y divided by sample size (n)
how to find median
value of Y (n+1)/2
how to compute 1st quartile
0.25xn
if an integer- avg position of j and j+1
if not - round up
how to compute 3rd quartile
same as 1st quartile in rules but instead 0.75xn
how to find interquartile range
q3-q1
how to calculate standard error
standard deviation/ square root of sample size
what is the problem with calculating standard error
-imaginary, not measured directly
how to find confidence intervals
LL= mean-2(SE)
UL=mean+2(SE)
what are the best error bars
those that do not use standard deviation- no precision
how to find cv percentage
standard deviation/meanx100
what is estimation/statistic
estimation of parameter based on sample
guidelines for effective graphs
labeled, simple colors and font, units, start at 0
