W7: Data Analysis I Flashcards
What are the steps of the research process?
- Identify topic of interest
- gather background info and ensure access
- set research question(s)
- choose design and methodology
- develop tools and methods
- access sample: begin data collection
- Organize and analyze data
- Interpret your results; define conclusions
- Transfer knowledge
What are the different types of statistics?
- Descriptive
- Inferential
Define descriptive stats. Give an example
procedures/tools for describing individual variables, relationships between variables
eg. describing characteristics of study sample (sex, age, etc.; often table 1 in articles)
Define inferential stats and give an example.
procedures used to analyze data after experiment is done & determine if an IV has significant effect; allow extrapolations from sample to population
eg. between group design
What are the 3 levels of measurement? (descriptive)
- nominal
-ordinal
-ratio
Define nominal and give an example.
numeric values are arbitrary, not ordered
eg. gender, political affiliation
Define ordinal and give an example.
has continuum, values are ordered but intervals are not equal.
eg. Likert items, income brackets
Define ratio and give an example.
numeric values assigned reflect equal intervals; has true zero
eg. weight, age in years
Define Tendency and list the types
simple stats that typify set of values
- mean
-median
-mode
When is mean used?
for describing ratio data
- aka arithmetic mean/average
When is median used?
for describing ratio and ordinal
When is mode used?
for describing ratio, ordinal, and nominal data
What are measures of dispersion and what are the types?
describes variability of measures
- range
- standard deviation
- variance
What is range? What kind of data is it used for?
calculated by subtracting lowest form highest value in set of values/ indicating lowest and highest values
(for ordinal and ratio data)
What is standard deviation (SD)? What kind of data is it used for? What is it related to?
measure reflecting the average amount of deviation from meal value in set of values
(ratio level data)
related to normal distribution curve
What is variance? How is it calculated? What kind of data is it used for?
single number that represents total amount of variation in distribution
variance = SD squared
(ratio level data)
What is the purpose of standardizing data? Give an example.
allow for comparisons between units of different sizes.
eg. suicide rates
9.9/100,000 - Ontario
32.4/100,000- Northwest territories
What are the 5 types of standardizing data?
- proportions
-percentages
-percentage change
-rates
-ratios
Define proportions. What can it substitute? How is it calculated?
standard method of designating portion of total; ranges from .00 (none of total) to 1.00 (all of total)
- can be used instead of percentages
- batting average (BA)
BA= # of hits/# at bats
what is a percentage?
proportion may be converted to % by multiplying by 100; denotes probability- how often something happens per 100 times
What is percentage change? How is it calculated?
measure of how much something has changed over given time period.
(time 2-time1/time 1) x 100
Define rates. Give an example.
represent frequency of something for a standard sized unit.
eg. divorce, suicide, crime
Define ratios
represent comparison of one thing to another
what is normal distribution?
aka normal curve/bell curve; based on concept: as number of scored in sample increases, scores approximate more so a “normal distribution”
half scores above and half below mean
What characterizes a normal distribution (ND)?
mean, median, mode are same
distances from meal of a normal distribution can be measured in SD units
2/3 of cases +/- SD of mean (by definition)
(68.2% of scored in ND)
Define Z-score
standardized score; represents distance above/below mean, in SD units, of any raw value in a distribution
usually range from +3 to -3 in value
What are the different types of relationship between variables?
- cross tabular analysis
- comparison of means
- correlations
What is cross tabular analysis. What kind of data is it used for?
used with nominal DV
data is cross- classified (sort into categories within IV and DV) to show relation between IV and DV
when is comparing means used? What is its purpose? What tests of significance may be used for comparison of means?
when DV is ratio level data and when IV is either nominal/ordinal
- compare mean values of DV for each category of IV
- t-tests and ANOVA (test of sig.)
When is correlation used (kind of data)? What is its purpose?
used to describe relationship between ratio level variables
describe how closely 2 ratio level variables co-vary together
(correlational analysis, correlational techniques)
What are the two basic questions asked of the correlation?
- what is the equation that describes relation between 2 variables?
- What is the strength of the relation between 2 variables?
What is the general equation for correlation?
Y= a +bX
estimates how much IV has to change in order to produce a unit of change in DV
What is a regression line?
line that offers best linear description of the relation between 2 variables
What is the correlation coefficient, r?
symbol used to report strength of correlation between 2 variables
can vary from -1.0 to +1.0
What is the meaning of r= -1.0?
perfect negative correlation- negative slope; increase of 1 unit in 1 variable is associated with proportional decrease in other variable
What is the meaning of r= +1.0
perfect positive correlation; positive slope