-
IV Examples
- Mood when manipulated by researcher
- Movie when selected by researcher.
-
True IV:
Quasi-IV:
Many quasi-IVs are....
Researcher sets or manipulates IV
Researcher observes BUT DOES NOT MANIPULATE level
participant variables
-
Quasi-IV
Researcher observes BUT DOES NOT MANIPULATE level
-
Example of Quasi-IV
- Mood when participant shows up for study
- Movie when selected by participant
- Sex of participant
-
Cell or Condition
Ex:
A combination of IV levels
Study of sex differences among drivers in two countries (USA & Greece)
-
Confounding variable
Other names =
What do they do?
- AKA confound, covariate, or 3rd variable
- Affects DV. but is not an IV
- Can bias study, lead to wrong conclusion
- Ultimately has affect on IV
(IV STIMULUS --> DV RESPONSE --> Confounding variable --> IV STIMULUS)
-
-
Whats the order
Date -> ______ ->
Data, statistical analysis, Information
-
What are the two types of statistics?
- Descriptive Statistics
- Inferential Statistics
-
Descriptive Statistics:
Goal?
Problems?
Best way to describe a variable is?
Summarize/Describe the data from your study (the sample)
Goal:Try to summarize data with two numbers without sacrificing crucial data:
- Average = Signal
- Variability = Noise
- Problems:
- Information loss
- Assumptions
Best way to describe a variable is by its type (nominal, ordinal, etc) and distribution
-
Inferential Statistics:
Are used to make predictions/inferences about the population based on the sample
-
Average:
Common averages:
Single value most typical of all data
- Mean =
- Meadian =
- Mode = Most freq value
-
Mode =
Works best for what distribution?
Used to describe?
Okay for what types of data?
- -Most freq value
- -Works best for unimodal distribution
- -Used to describe average of nominal variable
- -Okay for what types of data:
- (Ordinal Interval or Ratio)
-
Median =
APA format =?
- Apa = Mdn
- Divide Data in half ( 125 )
-
Mean =
Used to describe which data?
-
Mean = Best guess
average of interval or ratio variables, but only if distribution is unimodal and symmetric (not skewed)
-
IF distribution is unimodal and symmetric
Mode = Mdn = M
-
If distribution is unimodal with positive skew:
Mode < Mdn < M if unimodal
-
Would you report Mean, Median, or Mode?
Eye Color
Test Score (0-50)- Normal Distribution
Test score (0-50)- Highly Skewed Distribution
Temperature-- Normal Distribution
SIze of dog (small, medium, large)
Exam Grades (A, B, C, D, F)
INcome under (10k, 10-20k, 20-30 etc)
Eye ColorTest Score (0-50)- Normal DistributionTest score (0-50)- Highly Skewed DistributionTemperature-- Normal DistributionSIze of dog (small, medium, large)Exam Grades (A, B, C, D, F)INcome under (10k, 10-20k, 20-30 etc)
-
What is mean, median, mode of
1
1
4
-
Which would u use? Mean median mode
Small
Small
Small
Small
Medium
Medium
Medium
Large
Large
Mode????
-
Does average completely describe the data?
Solution?
No, because two distributions can have same average but look very different.
Need at least one other descriptive statistic.
-
Variation
Aka: Dispersion/Spread
*Single most typical difference from average
*Need to describe how much scores differ from average (ho wmuch noise is present)
-
IQR
Extends from?
Often used for?
Variabilities: How to measure them?
Range:
IQR:
- Interquartile Range- Middle half of scores
- 25th - 75th percentile
- Ordinal Data
- Variabilities:
- Range = Largest datum - smallest datum
- IQR= Q3-Q1
- where Q3 is the median of the upper half and Q1 is the median of the lower half of data. Q2 is the median of all data
-
Variance =
Squared Standard Deviation
-
Standard Deviation =
Know formula
Mean deviation from M, ignoring sign (sort of)
-
Standard Deviation
Different Mean =
Same Mean =
Different Mean =
- Different Mean = Same SD
- Same Mean = Different SD
- Different Mean = Different SD
-
Descriptive Statistics to report in Method Section/Participant subsection
- Sex: # of M & # of F
- Age: Min, Max, and either M and SD (of symmetric) or Mdn and IQR range (of skewed)
Other key demographics such as ethnicity if important
-
Descriptive Statistics of DV to report in...
- Results Section
- If DV is nominal or ordinal with few levels:
- Mode for each cell
- If DV is ordinal, or interval/ratio and skewed: n, Mdn, and IQR rang for each cell
If DV is interval/ratio and symmetric: n, M, and SD for each cell
-
Descriptives calculate:
M, SD, Min, and Max
|
|