Exam 1 Flashcards
(39 cards)
1
Q
Quantitative characteristics
A
- N = population size
- correlation doesn’t equal causation, correlation is an association
2
Q
Quantitative strengths
A
- good at finding correlation
- yields many responses (more representative)
- easier to chart
- gives a general outlook on a social situation
3
Q
Quantitative weaknesses
A
- not good at finding causation
4
Q
Qualitative characteristics
A
- n = sample size
5
Q
Qualitative strengths
A
- easier to establish causation
- in-depth
6
Q
Qualitative weaknessess
A
- generalization more difficult to establish
- not applicable to the general population
7
Q
Population
A
- total set of subjects of interest in a study
8
Q
Sample
A
- subset of the population on which the study collects data
9
Q
Parameter
A
- numerical summary of the population
- the value you are trying to uncover, cannot often do it precisely
- we do not always have access to the entire population
10
Q
Statistic
A
- numerical summary of the sample data
- to get a better sense of what the perimeter value might be
11
Q
Descriptive statistics
A
- statistics summarizing (outlining) sample or population data
12
Q
Inferential statistics
A
- statistics making predictions about population parameters based on sample data
13
Q
Qualitative variable
A
- variable that is placed on a measurement scale that has numerical values
14
Q
Quantitative variable
A
- variable that is placed on a measurement scale that has a set of categories
15
Q
Discrete variable
A
- variable taking the form of a set of separate numbers, such as 0, 1, 2, 3
16
Q
Continuous variable
A
- variable that can take an infinite continuum of real number values
17
Q
Measurement scales: interval
A
- quantitative, scale with specific numerical distances between levels
- Nominal: qualitative, scale with categories that are in no specific order
- Ordinal: qualitative, scale with categories that are in a specific order
18
Q
Sampling methods
A
- Random sampling: drawing a sample of n subjects who each have the same probability of being drawn, ability to better make inferences / draw conclusions about the population
19
Q
Sampling errors
A
- Sample bias: sample is not representative
- Response bias: under and over reporting
- Non-response bias: large sample, few participants
20
Q
Distribution shapes
A
- normal (bell-shaped)
- U-shaped
- skewed to the right
- skewed to the left
21
Q
Frequency
A
- relative frequency: proportion of observations falling into a category
22
Q
Frequency distributions
A
- bar graph (typically categorical data)
- pie chart (typically categorical data)
- histogram (typically quantitative data)
23
Q
Central tendency: mean
A
- an average
- sum of the observations divided by total number of the observations
24
Q
Central tendency: median
A
- observation that falls in the middle of the ordered sample
- what is the most typical observation you can come across
- mean and median are usually close in a normal distribution
25
Central tendency: mode
- value that occurs most frequently in the distribution
- unimodal: for one mode
- bimodal: for two modes
- multimodal: for more than two modes
26
Variability: range
- difference between the largest and smallest observations
27
Variability: deviation
- difference between an observation and the mean (i.e. how far an observation ‘falls’ from the mean of the population or the sample)
28
Variability: standard deviation
- typical (average) deviation from the mean for an observation in the set
- will always have a positive value, but can go either way
29
Variance
- approximate average of squared deviations in a distribution
30
Percentile
- measure of data dispersion breaking down the distribution in percentage points.
- an observation’s percentile indicates the percentage of observations that are of equal or lesser value in the distribution.
- conversely, an observation’s percentile also allows us to calculate the percentage of observations that fall above it in the distribution.
- impossible to have a 100 percentile (implies that 100% of the population has the same value or be below your value)
31
Quartile
- measure of data dispersion breaking down the distribution in four ordered segments.
- when ordered in ascendance, the first 25% of the data distribution comprise the lower quartile, whereas the first 75% of the distribution comprise the upper quartile
32
Quartile order
- Min: 0% of the data
- Q1: First 25% of the data
- Med: First 50% of the data
- Q3: First 75% of the data
- Max: Fully 100% of the data
33
Empirical Rule
- ~ 68% of observations fall between the mean and one standard deviation on either side
- about 2/3 will fall on both side of the middle
~ 95% of observations fall between the mean and two standard deviations on either side
- two standard deviations (20-7-7 — 13, 6)(95% will be greater than 6 and lower than 13)
- over 99% of observations fall between the mean and three standard deviations on either side.
- data will be cluster around the middle
- rare to find observations that fall off of the distribution
34
Z-score
- the number of standard deviations that any given observation in a distribution falls away from the mean of that distribution
- tells you the right tail probability associated with that z-score, can also use it to find the LTP
- RTP is the probability of encountering another observation that is further away removed from the mean than the observation in question
35
Sampling distribution of a statistic
- probability distribution that specifies probabilities for the possible values the statistic can take
- every sample a mean; can draw a distribution form the mean
36
Sampling distribution of sample means
- the probabilities of specific values the mean of a sample would take if we repeatedly drew random samples from the population
- the sample distribution of y-bars (distribution of the means of the samples that we have collected)
- whereas a ‘regular’ probability distribution has a standard deviation (σ), a sampling distribution of sample means has a standard error (σȳ).
- same concept of a standard deviation
37
Central Limit Theorem
- for random sampling with a large sample size n, the sampling distribution of the sample mean ȳ is approximately a normal distribution (n=30 is sufficient)
38
Measurement scales: nominal
- qualitative, scale with categories that are in no specific order
39
Measurement scales: ordinal
- qualitative, scale with categories that are in a specific order
