types of variables

what is the hierarchy of evidence?

- a ranking of the quality and reliability of research studies with strongest evidence at the top

what is the order of the hierarchy of evidence from top to bottom

- Randomized control trial

- cohort study

- case control study

- cross sectional study

- ecological study

- case series

what is a cohort study

- follow people over time to see who gets a disease

what is a case control study

- compare people with and without disease

what is an ecological study

- comparing characteristics of groups (like geographical comparisins)

what is a case series study

- clinical observations

what is a target population, sample frame, and sample?

- target population is the population of interest that the researcher wants to generalize the study to

- the sampling frame is the part of the target population that is accessible for the study

- the sample is the actual units selected for the study and should be as representative of the target population as possible

explain an independent variable

- an independant variable is changed or controlled to test the effects on the dependent variable

- its value is not affected by other variables in the study

explain a dependent variable

- the variable that is being measured

- it is expected to change in response to the independent variable

- outcome that the researchers are interested in observing

name the indemendent and dependent variable in this example: Is the number of Home exersice sessions completed per week associated with pain levels in people with chronic low back pain?

- Independent: HES/week

- Dependent: pain levels

explain nominal values

- name variables with no value attatched to them. they are used to classify things

- there is no ability to rank or order these variables

- Ex: blood type, sex, race, gender

what are the statistics allowed for nominal values

- count only

- can compare proportions in categories

- identify differences in proportions using chi-square analyses

explain ordinal measurements

- classification variables that can be placed in an order

- pain scale, poor, great, good, levels in education

statistics allowed for ordinal measurements

- counts

- use numbers to label but not to do math with (symbols and not numerical quantities)

- calculate percentages and proportions

- can do rankings to determine significance

- can do chi square analysis

what is an interval measurement

- number variables in which the order and intervals between the variables are known

- different than other number variables because the origin is unknown and there is no true 0 point

- ex: temperature, pH, SAT score, credit score, scaled measurement tools like pain scales or addiction scales

statistics allowed for interval measurements

- mathematical calculations are allowed with some caution with more subjective non measureable items like pain scales

what are ratio measurements

- numbers where the order, interval distance and origin are all known

- ex: weight, height, number of children, lab values (blood pressure, cholesterol)

stats allowed for ratio measurements

- mean

- median

- mode

- all statistical manipulations

what are the two types of numerical data

- interval and ratio

- can be discrete or continuous

what is discrete data

- can only take on certain numeric values

- often whole numbers

- ex: number of births, number of survivors on the titanic

what is continuous data

- not restricted to whole number values

- often measurements

- ex: height, weight, age

what is time to event

- time between initial and terminal events

- ex: diagnosis of breast cancer to cured

- some subjects are censored

explain why subjects are censored

- they never experience the terminal event during the study

- they could be lost to follow up, experience death from the condition, death from another cause

what are descriptive summaries

- allow us to identify patterns within data and understand relationships with variables

- check hypothesis

- formal hypothesis testing should agree with descriptive results

what are the types of descriptive statistics?

mean, median, mode, percentile

what is the mean

average value

- add up the numbers and divide by the number of numbers

what is the median

- the middle number when arranged from least to greatest

- the 50th percentile where 50% lies above the median and 50% below

what is the mode

most frequently occurring value

what is the percentile

- the value below a given percentage

- ex: if you are in the 90th% it means you scored better than 90% of the people

data that are not symmetric and bell shaped are....

skewed

if the median and the mean value are not very similar the data could be...

skewed

a data set where the mean > median is...

positively skewed

a data set where where the mean< median is...

negatively skewed

how do we measure spread or variation

variance: average squared deviation from the mean

standard deviation: square root of variance

the more spread out the data is .......

the more variation and more difficult to measure

what are barplots used for

- to describe distribution of values for categorical variables and a single variable

- determine categories and draw a bar with height equal to proportion of responses in each category

what are histograms used for

- to describe the distribution of values for a continuous variable

- determine intervals of response and draw a bar with height equal to proportion of responses

what is a boxplot used for

- describe distribution of values for continuous variables

- determine the 25th, 50th, and 75th percentiles of distribution

- determine outliers

- draw a box with lower line at 25th middle line at 50th and upper line at 75th

- add whiskers for minimum and maximum values

what is a scatter plot used for

- describe joint distribution of values from two continuous variables

- plot pair of observations for each subject from both variables

should scatter plots be reported with or without outliers?

-both so as to not bias the data

what is a survival curve

- distribution of times to an event

- ex: time to heal, time to death, time till ready for prosthesis

- estimate survival probabilities while accounting for censoring, plot the survival probability corresponding to each time an event occurred

- axis are often survival probability and time in years of observation