Lecture 1 Flashcards
Introduction to Statistics
What’s the framework for design and statistics?
- Hypothesis/question
- Propose a study
- Design the study, i.e., how we get the data
- Collection of data
- Use of statistics to test hypothesis on model of data.
- Examine and interpret the results
Variable: Infinite Number of possible values, i.e., entities get a distinct score
A. Continuous variable
B. Categorical variable
A. Continuous variable
Variable: Cannot take on all values within limits of the variable, i.e., entities are divided into distinct categories
A. Continuous variable
B. Categorical variable
B. Categorical variable
Equal intervals on the variable represent equal differences in the property being measured.
A: Interval
B: Ratio
C: Nominal
D: Ordinal
A: Interval
Example: Difference between reaction time
Variable has a clear definition of 0:0
A: Interval
B: Ratio
C: Nominal
D: Ordinal
B. Ratio
Example: Height
Variable: Two or more categories.
A: Interval
B: Ratio
C: Nominal
D: Ordinal
C. Nominal
Example: One person one outcome, vegetarian/vegan/nonveg
Variable: Categories have logical, incremental order
A: Interval
B: Ratio
C: Nominal
D: Ordinal
D. Ordinal
Example: Likert Scale, fail/pass/merit, ranked
What’s the measurement area?
A: Validity (instrument measures what it set out to measure)
B: Reliability (ability of measure to produce same results under same conditions)
C: Both A and B
D: None of the above
C. Both A and B
The values have to have the same meaning over time and across situations, and must have systematic issues with measurements
Variation: Differences in performance created by a specific experimental manipulation
A: Systematic variation
B: Unsystematic variation
C: Randomization variation
D: Other/NOTA
A. Systematic variation
Influencing the outcome based on a controlled experiment
Variation: Differences in performance created by unknown factors
A: Systematic variation
B: Unsystematic variation
C: Randomization variation
D: Other/NOTA
B. Unsystematic variation
It influence results, but it’s not accounted for.
Example: Age, gender, IQ
Variation: Minimizes unsystematic variation
A. Systematic variation
B: Unsystematic variation
C: Randomization variation
D: Other/NOTA
C. Randomization variation
It minimizes the impact of unsystematic variation
What is the independent variable?
A. Predictor variable
B. Outcome variable
C. Both A and B
D. Neither
A. Predictor Variable
Also referred to as Hypothesized cause, manipulated variable.
What is the dependent variable?
- Measured variable
- Manipulated variable
- Both A and B
- Neither
- Measured variable
Also called outcome variable, and is the proposed effect.
NHST
Which hypothesis predicts no effect of predictor variable on outcome variable?
- Null hypothesis
- Alternative hypothesis
- Both 1 and 2
- Neither
- Null hypothesis
NHST
Which hypothis says that there is an effect of the predictor variable on outcome variable?
- Null hypothesis
- Alternative hypothesis
- Both 1 and 2
- Neither
- Alternative hypothesis
NHST
What does NHST compute?
1. Probablity of null hypothesis being true
2. Probability of alternative hypothesis being true
3. Probablity of null hypothesis being false
4. None of the above
- Probablity of null hypothesis being true
Also called P-value
NHST
Which of the following is true?
1. Significant result does not mean effect is important.
2. Non significant result does not mean null hypothesis is true
3. Significant result doesn’t mean that the null hypothesis is false.
4. All of the above
- All of the above
NHST
Which of these refers to hypothesizing after the results are known?
1. p-hacking
2. harking
3. other
- Harking
NHST
What is p-hacking?
Selective reporting of significant results
EMBERS
What is E?
Effect size
It is the standardized measure of how different the means are.
Example: Cohen’s d
EMBERS
M - ?
Meta analysis
Rule of thumb for effect sizes
Funnel plots: value studies by their sample size and observe bias.
EMBeRS
What is BeRS?
Baynesian approach, Registration, Sense
Distribution
What is a normal distribution?
- Bell curve
- Symmetrical
- Two parameters
- Central tendency (mean)
- Standard deviation (dispersion)
Symmetry and Skew
If mean = median = mode, then:
1. Symmetrical
2. Positive skew
3. Negative skew
- Symmetrical
Skew is between -1 and 1
Symmetry and Skew
Is mean > median > mode, then
1. Symmetrical distribution
2. Positive skew
3. Negative ske
- Positive skew
x > +1
Symmetry and Skew
Is mode > median > mean, then
1. Symmetrical distribution
2. Positive skew
3. Negative skew
- Negative skew
x < -1
Kurtosis
What does kurtosis look at?
- Skewness
- Shape of graph
- Length of graph
- Direction of graph
- Shape of graph
Bulge or bend.
If within +2 and -2, its all good!
Kurtosis
Which is leptokurtosis?
1. k > 0
2. k < 0
3. k > +2
4. k < -2
- k > 0
Kurtosis
Which is platokurtosis?
1. k > 0
2. k < 0
3. k > +2
4. k < -2
- k < 0
Tortium Quid/Third factor
Which is NOT the assumptions of tortium quid do?
1. It helps infer causality
2. It has randomized controlled trials
3. It acts as a confounding variable/third factor
4. All of the above
- All of the above
It creates control settings to see how the 3rd controlled variable affects the data