Test 1 Flashcards
Frequency
of times occurred
Relative frequency
occurrence proportional to sample size
Relative percent
relative frequency x 100
Cumulative frequency
of scores equal to or below the score
Cumulative relative frequency
cumulative frequency proportional to sample size
Cumulative percent
cumulative relative frequency x 100
Polygons
graph with connected dots (both ends come down to 0)
Histograms
Graph with adjoining bars
Bar graph
graph with separate bars- used for nominal level data
Levels of Measurement
- Nominal
- Ordinal
- Interval
- Ratio
Nominal
No properties of measurement
Use #s to categorize
Ex: gender, ethnicity, helped or did not help
Ordinal
Magnitude but no equal intervals or absolute zero
Often rank ordering
Ex: places in a race
Interval
Magnitude and equal intervals but no absolute zero
Provides info about the amount of difference between scores
Ex: degrees Fahrenheit
Required to use preferred statistics
Researchers often treat measurements that are technically ordinal as if they were interval (ex: Likert scales)
Ratio
Magnitude, equal intervals, and absolute zero
Ex: length, duration
Allows for ration comparison
Which level of measurement is most informative?
Ratio scale
Discrete vs. continuous variables
-Discrete: Measured in whole units or categories that are not distributed along a continuum
Theoretically a finite # of possible values between any two values (Ex: # of siblings, # of points scored in basketball)
-Continuous: Measured along a continuum at any place beyond the decimal point
Can be measured in fractional units
Theoretically an infinite # of possible values between any two values (on a continuum) (Ex: blood pressure, intelligence)
Experiment
Best method to demonstrate cause and effect
Researcher tries to ensure that all other possible causes, except for the effect of the IV, are unlimited or highly unlikely
Strategy: while controlling extraneous variables (anything that affects DV), manipulate an IV, and then measure effects on DV
To control for extraneous variables (hold constant, balance; ex: RA)
Goal: to isolate cause and effect- participants’ experience in the research is the same except for manipulations of IV
Quasi- experiment
Like an experiment, but lacks at least one characteristic of a “true” experiment
Could lack:
A manipulated IV (ex: compare pre-existing groups)
A comparison/ control group
Correlational study
Measure pairs of scores
Lacks controls needed to determine cause and effect~ 3 conditions
1. Covariation (only one met with correlational)
2. Time order~ cause must precede effect
3. Control of extraneous variables
Examine the extent to which two variables change with one another
Descriptive vs. Inferential Statistics
Descriptive: Evaluation of info where statistics organize and summarize info such that the info is meaningful to those who read about the observations scientists made in study (to organize/ summarize info)
-Data you have
Inferential: Evaluation of info where scientists use info to answer a question or make an actionable decision (to interpret meaning of info)
-Beyond your data
Frequency table when there is grouped data
Use grouped data b/c much data with a wide range of scores
5 to 20 intervals ( <5 then over-summarize, >20 then overly detailed)
To construct frequency table for grouped data
- Find real range: highest score - lowest score + 1 (round if possible to a whole number)
- Interval width: real range / desired # of intervals
- Construct frequency distribution
-1. Lower and upper bound
-2. Intervals should be = in size
-3. No interval overlaps
-4. Values rounded (to the same as original data)
-5. Order intervals from high to low
-6. Lower bound of lowest interval needs to be as low as
(or lower) than the lowest score
APA style table
3 overarching goals:
- Simple
- Complete
- Self-explanatory
