Midterm Flashcards
1
Q
What is falsification?
A
The act of disproving a theory or hypothesis.
2
Q
What is an independent variable?
A
- The proposed cause.
- A predictor variable or a manipulated variable.
3
Q
What is a dependent variable?
A
- The proposed effect.
- An outcome variable that is measured and not manipulated.
4
Q
What are the 3 categorical variables?
A
- Binary (only two categories)
- Nominal (there are more than two categories)
- Ordinal (these same as nominal but they have a logical order. E.g., first place, second place, third place)
5
Q
What are the 2 continuous variables?
A
- Interval (equal intervals on the variable represent equal differences in the property being measured. E.g., a scale of 1-10 or temperature)
- Ratio (the same as interval, but we have a “true” zero. So weight is an example)
6
Q
What is measurement error?
A
- The error between the actual value we’re trying to measure and the number we use to represent the value
- Example: your dog weighs 40 pounds, the scale says 38 pounds. The measurement error is 2 pounds.
7
Q
What is correlational research?
A
- Observing what naturally goes on in the world without directly interfering with it.
8
Q
What is cross-sectional research?
A
- This term implies that data come from people at different age points with different people representing each age point.
9
Q
What is experimental research?
A
- One or more variable is systemically manipulated to see their effect on an outcome variable.
10
Q
What are the 2 methods of data collection?
A
- Between group (different entities in experimental conditions)
- Within group (the same entities take part in all experimental conditions)
11
Q
What is systemic variation?
A
- Differences in performance created by specific experiments manipulation
- it is what we are interested in
12
Q
What is unsystemic variation?
A
- Differences in performance created by unknown factors
-Age, gender, IQ, time of day
13
Q
What can minimize unsystematic variation?
A
-Randomization
14
Q
What does a normal distribution look like on a histogram?
A
- Bell shaped curve
- Symmetrical around the centre
- You expect this most with: height, weight, age, IQ
15
Q
What are the two properties of frequency distributions?
A
- Skew (the symmetry of the distribution)
- Kurtosis (looking if it is peaked or flat)
16
Q
What are positive and negative skews?
A
Positive - scores bunched at low values with the tail pointing to high values
Negative - scores bunched at the high values with the tail pointing to low values
17
Q
What is leptokurtic?
A
In kurtosis, it is heavy tails (tall)
18
Q
What is platykurtic?
A
In kurtosis, it is light tails (short)
19
Q
What is the mode?
A
- The most frequent score
20
Q
Can a normal distribution have more then 1 mode?
A
- No a normal distribution only has 1 mode (1 peak on the graph)
21
Q
What is the median?
A
- The middle score
22
Q
What is the mean?
A
- The average
23
Q
What is the range?
A
- The smallest score subtracted from the largest score
- It is very biased by outliers
24
Q
What are z-scores?
A
- It expresses a score in terms of how many standard deviations it is away from the mean
- “On average how much do people differ from the average”
- The distribution of z-scores has a mean of 0 and SD=1.
25
What are the 2 types of hypotheses?
- Null hypothesis (there is no effect)
- Alternate hypothesis (there is an effect)
26
What does this formula tell us
Outcome = (model) + error
Model is the data we are calculating
Error is the standard deviation (could be a measurement error, sampling error, etc.)
27
What is a simple statistical model?
- A model to represent what is happening in the real world
28
How do we calculate error?
- It is the deviation score
- It tells us how much everyone differs from the average
- The smaller the deviation the better our model fits or represents our data
29
How do we get the sum of errors?
- we take the error between the mean and the data and add them
30
How do we find the sum of squared errors?
- we square each deviation and then add them
31
How do we find variance?
- Take the sum of squared errors and divide it by the sample size minus 1
32
How do we take the average error and make it a standard deviation?
- We scare root the value
33
If our mean was 2.6 hours of studying per week and our standard deviation is 1.14 what does this mean?
- On average, our sample varied 1.14 hours of studying
34
How much % of data fall within the standard deviation
- 68% of data are within +/- 1 standard deviation of the mean
35
How many deviations is it before we reach 99% of our sample
- +/- 3
- We do not want to exceed this
36
What does a standard deviation of 0 mean?
Means there is no difference
37
Why do we subtract 1 in the standard deviation when calculating it?
Because of the degree of freedom
38
Example of degree of freedom
If you know in the population the mean is 10.
You know 4 out of 5 scores.
Then what does the 5th score have to be so that the mean is 10
It can only be a specific value to make that average a 10
That would be your degree of freedom
39
You want to see whether there is a difference in literacy based on an after school reading program.
What is the null and alternate hypothesis?
Null - there is no difference in literacy after an afterschool program
Alternate - There is a difference in literacy after an after school program
40
Outcome = (model) + error
We are examining the number of hours playing sports among children, what would our model be?
- Average hours spent playing sports
41
The mean hours of sports per week in our sample is 5.3.
What does this mean in plain language?
- The average number of hours playing sports per week is 5.3
42
The mean hours of sports per week in our sample is 5.3
We have a standard deviation of 2.2 hours
What does this standard deviation mean in plain language?
- the group varies by about 2.2 hours
43
How many groups are involved in a one sample t test?
1 sample (testing 2 groups)
44
Who are we comparing our sample to in a one sample t test?
Known population
45
How many groups are involved in a paired samples t test?
2
46
Are the groups in a paired samples t test the same people?
YES (Same people in each group, first measure before manipulation, second measure after manipulation)
47
What are the two groups in a paired samples t test called?
Pre group and post group
48
Can we reject a null?
Yes
49
Can we accept a null?
No
We can’t know for sure
We can fail to reject the null
50
Why is the alpha set at p = .05
Rate of finding an effect when the null is true, want less then 5% error rate
51
To meet our assumption we want which direction < >
We want p > .05
52
For statistical testing and discovering something cool we want to
p?
We want p < 0.05
53
Why do we need continuous data?
Needs to meet assumption
Normality not relevant to categorical
54
What are ways to assess normality?
Skewness/kurtosis values
Histogram
Probability/quantile plot
KS test
Box plot
55
When do we check independence?
During the design stage
56
How do we check homogeneity of variance?
Through the spread of scores
57
How do we check for outliers?
Box plot
Standardized scores (+/- 3 SD)
58
What can you do to remove outliers?
Remove
Transform
Change
(But you need to justify)
59
You have conducted a KS test of normality on a variable called “social” assessing students social and emotional skills. The p value is 0.038. What is the null and alternate hypothesis?
Null- social and emotional skills variable is not significantly different from a normal distribution
Alternate- social and emotional skills variable is significantly different from a normal distribution
60
You have conducted a KS test of normality on a variable called “social” assessing students social and emotional skills. The p value is 0.038. What would you conclude?
Step 1: what is the decision rule
- Is p < 0.05? YES
Step 2: do we reject the null?
-YES
Which means the variable is not evenly distributed
61
What is least affected by outliers?
The median
