Surveys, Samples, Experiments, Data Displays, Descriptive Statistics

When skewness is present in a set of data, which descriptive summary measure is most appropriate

IQR and median

if high concentration of data in the middle...

IQR is small

volunteer/self-selected bias

a call goes out and people enter the study on their own - no sampling procedure and does not represent any population

undercoverage bias

a subgroup of the population is excluded from the very beginning - sampling procedure is used but can only represent the remaining population without the subgroup

response bias

an individual in the sample responds but does not give the correct data

nonresponse bias

an individual is selected to be in the sample but does not respond to the survey - look for high percentage not number

statistical significance

any sizeable enough difference is deemed to be due to treatment - difference beyond chance - due to more than random chance

sample mean

average (x̅) - does not have to be one of the numbers in the data set

convenience bias

choose individuals in the easiest way - sampling procedure but does not represent any population

stratified random sample

compare subgroups of the population equally - divide population into subgroups - choose a simple random sample from each subgroup

a flat histogram...

contains more variability

categorical data

data falling into groups

random sample

each group of the same size has the same chance of being selected as the sample

simple random sample

examine the entire population as it exists

independent variable

factor, x

if the mean of a data set is large, the standard deviation has to be large also

false

control group

given fake or no treatment, placebo

experiment

imposes some treatment and observes their responses

data distribution

list of all possible outcomes and how often they occur

if data is symmetric...

mean and median are similar

skewed right

mean is greater than median

skewed left

mean is less than median

if there are very few large values in a data set compared to the rest of the data...

mean will be greater than median

if there are very few small values in a data set compared to the rest of the data...

mean will be less than median

quantitative data

measurements and counts - numbers, money

from boxplot, can see...

median, IQR, and whether or not data is skewed

five number summary

minimum, Q1, median (Q2), Q3, maximum

treatment group

set of subjects given a certain treatment

which is more affected by skewness, IQR or standard deviation

standard deviation

...can affect graph

starting point and number of bars

most common observational study

survey

bias

systematic favoritism

timing

timing of survey can affect results

an outlier in a data set can significantly affect the value of the mean but not the median

true

you can have two data sets with the same mean but different standard deviation

true

type of survey

type of survey conducted can affect the results

median is least affected when...

values are added to the data set

confounding variable

variable not studied that can affect results

spot/avoid problems in bar graphs

watch the scale and look for sample size

histogram

way to graph quantitative data

if you add 10 to every value of a data set, standard deviation...

will be the same

if you multiply every value of a data set by 10, standard deviation...

will change

question wording

wording of a survey question can affect the results

standard deviation of the data set 1, 1, 1, 1

0

which set of four numbers would give you the largest standard deviation

1, 1, 4, 4

spot/avoid bias

1. sampling procedure must be used 2. sample must represent the entire population

Q1

1st quartile, 25th percentile

Q2

2nd quartile, 50th percentile, median

Q3

3rd quartile, 75th percentile

boxplots...

- cannot tell what sample size is - bigger boxes do not mean more data - cannot see mean

interquartile range (IQR)

- distance taken up by the middle 50% of the data - Q3 - Q1

good experiments...

- make comparisons - avoids bias - has enough data (random data)

standard deviation

- use to measure concentration of data around the mean - same units as original data - never negative - can equal zero - affected by outliers and skewness

frequencies

number in each category - table, bar graph with #

median

number that splits the ordered data in half - does not have to be one of the numbers in the data set

observational study

observes individuals, measures variables and makes conclusions or comparisons, and does not attempt to influence responses

relative frequencies

percentage in each category - table, pie chart, bar graph with %

in general, from a histogram, you cannot...

recreate the original data values

impact of bias

reduces credibility of results

dependent variable

response, y