Variables & descriptive stats Flashcards
Descriptive stats? (def, includes what things)
- Summarize or describe data from a particular sample
- Includes…
–> CT: mean, median, mode
–> SD: s, s hat
Inferential Stats? (def, includes what things)
- Infers from the data to draw a conclusion about the pop
- Includes…
–> T-test
–> Df
–> P
Characteristics of a true experiment?
- IV was actually manipulated by the researchers
- DV gets measured
- Usually involves random assignment
Between-groups design meaning? Ex?
- Participants only see ONE condition
- diff people in conditions, compare between them
- EX –> drug studies
Within-groups design meaning? Ex?
- all participants see BOTH conditions
- same participants in both conditions, make comparisons WITHIN the group
- EX. –> sleep study with cued & non-cued questions
Factors that effect between/within groups design?
Between:
- both conditions would give away the goal / design of study
- both conditions is impractical / impossible
Within:
- advantage b/c participants are their own comparison group
Operational definition? (def & ex)
- how the researcher chooses to measure or manipulate variables of interest
- Ex. they define the DV of “wellbeing” as the avg score on 2 different questionnaires
Discrete def? Ex?
- Specific values
- Whole numbers
- Can’t have partial units
Ex:
- shoe size, number of people, number of legs
Continuous def? Ex?
- full range of values
- includes decimals
Ex:
- height, weight exam scores with partial credit
4 diff levels of measurement? (in order)
- Nominal
- Ordinal
- Scale
–> Interval
–> Ratio
Nominal: memory cue, definition, Cont vs Disc, CT measures, & examples
- Nom = NAME
- Def: measured category or name
- Always discrete
- ONLY mode
- Fav color, major, etc
Ordinal: memory cue, definition, Cont vs Disc, CT measures, & examples
- Ord = ORDER
- ordered descriptions, rankings
–> categories, with some inherent order, but not evenly spaced - Always discrete
- Mode or median (never mean)
- 1st/2nd/3rd, Yes/maybe/no
Interval: memory cue, definition, Cont vs Disc, CT measures, & examples
- numbers are evenly spaced, but no meaningful zero
- often continuous
- M, median, mode –> all good
- IQ tests, Celsius/Fahrenheit, strongly agree/disagree questions
Ratio: memory cue, definition, Cont vs Disc, CT measures, & examples
- ratiO –> 0
- numbers equally spaced, have a meaningful zero point
- often continuous
- M, median, mode –> all good
- reaction time, counting
Scale def? CT measures?
Includes interval and ratio level of measurement
M, med, mode –> all good
Mean: what is it & what is it appropriate for (variable type and distribution type/ skew/ outliers)
- average –> (sum of x)/n
- use for scale variables
- NOT great for skewed distributions or many outliers
Median: what is it & what is it appropriate for (variable type and distribution type/ skew/ outliers)
- middle (50th %ile) of the data
- (N+1) / 2
- okay for ordinal & scale variables
- okay for any distribution type
Mode: what is it & what is it appropriate for (variable type and distribution type/ skew/ outliers)
- score that occurs most often
- okay for all variable & distribution types
Range: what is it?
- max - min scores
IQR: what is it? how to calculate it? comparison to range & SD
- contains middle 50%
- 75%ile - 25%ile
- better than range, worse than SD
Variance: what is it? how to calculate it? units?
- sum of squared deviations from the mean divided by the number of scores
- Σ(x-M)^2 / N
- Units: squared units
lowkey useless
Standard deviation: what is it? how to calculate it? what version does jasp give?
(don’t talk about all the diff symbols yet)
- variation from the mean in standard units
- sq root of Σ(x-M)^2 / N
- JASP gives est. pop SD always
Sample standard deviation is NOT a good estimate for population. It is always too ____ and is strongly effected by ____.
It is always too small
It is strongly effected by sample size
Standard deviation: symbols & what they represent? 3 categories
POP standard deviation: σ
Sample standard dev:
–> SD or S (big “S”)
Est. pop standard dev:
–> s, ŝ (little “s”)
Estimated pop SD? Symbol? Why? How to calculate it?
- symbol: ŝ (sometimes little “s”)
- Sample SD is always too small
- Affected by sample size
- N-1 in denominator fixes it
Confidence interval / interval estimate def? Descriptive or inferential??
- range of plausible values for pop. parameter centered around the point estimate (usually mean)
- inferential!! allow you to infer about the population
Why is the point est. in the middle of the CI?
(hint: how is it calculated?)
- the CI is constructed by adding and subtracting the margin of error to the point estimate
What does a narrow vs wide CI tell us?
- That we are more/less precise or sure of our point estimate
- Narrower is much better
How to tell if population mean estimates are pretty similar or not based on their CI?
Effect size: what’s it called? what does it measure, what are the units? what if it’s 0?
- Cohen’s d
- measures the difference between means of 2 groups
- units: standard deviation units
- d=0 if means are the SAME
What does effect size (d) tell you about the groups?
- the amount of overlap between the distributions
- how much the change in level of IV actually affected the DV you’re measuring