W10 Data Analysis II asynch. Flashcards
Define Inferential statistics
- next step after you have collected and summarized data
- used to make inferences from a smaller group of data to a possibly larger one
what is a population?
a complete collection of all elements (scores, people, etc.)
what is a sample?
A subset of elements drawn from a population
How can you alleviate the problem of needing a large sample?
By selecting samples and extrapolating what was found
What are we looking for in a representative sample?
- same sampling distribution (b/w sample and population of interest)
- similar sample mean and population mean
- similar measures of central tendency and dispersion (b/w sample and population)
what is the goal of using a sample?
to generalize some parameters for a population based on what is learned from the sample
what is an univariante analysis?
analysis of one variable….duh
what is a bivariante analysis?
- describes relationship/association b/w two variables
- test relationship b/w dependent and independent variable
- how independent influence dependent variable
- anaylsis of two variables…duh
what are the common types of bivariante analysis ?
- scatter plots
- correlation coefficient
- regression anaylsis
define correlation, what is another name for it?
- bivariante anaylsis that measures strength of association b/w 2 variables and direction of relationship
- bivariante correlation
what is a pearson correlation coefficient?
a numerical index that reflects the relationship b/w 2 continuous variables
what are the directions of a relationship?
- positive (+ correlation coefficient)
- negative (- “ “)
what is the strength of relationship?
the value of the correlation coefficient
- ranges b/w -1 to +1
- used to describe strength of relationship
(higher value = stronger relationship)
What is regression analysis? What is it also known as?
aka the line of best fit
- minimizes distance b/w each individual point and the regression line
- enables predictions- how well x can predict y through the creation of the regression line
picking the right statistical test depends on…
nature of the question being asked and form of the null hypothesis
- Example: do you want to look at the difference between 2 groups or the relationship between 2
variables? - these 2 situations will result in different statistical approaches, but both will result in test of
null hypothesis using specific test of statistical significance
identifying the most appropriate statistical method is…
dependent on number of variables and type of data involved
- # of variables: univariate, bivariante, multivariante
- types of data: continuous, categorical
what is a chi- square? give an example.
2 categorical variables (each 2 or more levels)
- Eg. are hypertension and obesity status independent of one another?
what is logistic regression? Give an example.
1 categorical DV(BP) (only 2 levels)
1 or more categorical IVs
- Eg. what are the odds of developing hypertension (yes/no) for each race after controlling for gender?
Define confidence interval. What is it used to express?
- communicates how accurate our estimate of the outcome/parameter is likely to be
- range researchers are sure that true parameter lies somewhere b/w that range
- if experiment conducted over again, estimate expected to be within this range
The higher the confidence level, the greater…?
they are that relationship b/w variables (bivariante/multivariante) is not due to chance or other reason
Researchers need to pick a confidence level that will describe…?
how certain they are that the estimate will fall within a selected range
Define probability of error
-always chance of error
-risk associated with not being 100% correct
95% confident is association
5% experiment due to chance
Define P-value
probability that a test statistic would be as extreme as or more extreme than observed if null hypothesis were true
