-
Authority
someone says it's true
-
Rationalism
use of reasoning to arrive at knowledge (syllogisms)
-
Intuition
Sudden insights
-
Sample
group of participants selected for study
-
Population
the targeted audience of your research
-
Independent Variable
- manipulated by the experimenter
- -effect of alcohol on social behavior: IV=alcohol
- -sleep deprivation on academic performance: IV=sleep deprivation
-
Qualitative vs. Quantitative
- Qualitative- involves the type of treatment
- Quantitative- defined by the amount of a variable (dosage of drug)
-
Quasi-Independent Variable
Independent variables that are predetermined-making it impossible to randomly assign subjects
-
Dependent Variable
measured (measure of performance)
-
Reliability vs. Validity
- Reliability- consistency
- Validity- are you testing what you say you're testing
-
Descriptive vs. Inferential Statistics
- Descriptive- analysis done for purpose of describing or characterizing data (mean, mode, SD, variance)
- Inferential- using data for sample to make inferences or draw conclusions about the population
-
Internal vs. External Validity
- Internal- degree to which the researcher can infer that a relationship between the IV and DV is causal
- External- addresses the question of generalizability
-
Assumptions of Statistical Tests
- 1. Normality (normal distribution)
- 2. Homogeneity of Variances (variability between scores is roughly equivalent)
- 3. Independence of Errors (how the test is administered)
-
Nominal
- qualitative variables, names
- objects are measured by determining the category to which they belong
- Can compare modes but no means or medians
- -Social security number, sex of a child, religion of an individual
-
Ordinal
- Ranking, or magnitude (1st, 2nd, 3rd..)
- No equal intervals or absolute zero
- -Ratings of a task (easy, mildly difficult, difficult), marathon runners
-
Interval
- Magnitude and equal intervals between units, no absolute zero
- -Celsius/Fahrenheit Scales
-
Ratio
- Magnitude, equal intervals, AND absolute zero
- Allows for all types of numerical operations
- -time required to complete a task, processing speed, length, weight
-
Continuous vs. Discrete Variables
- Continuous- infinite number of values between adjacent units on a scale (weight, height, processing speed)
- Discrete-no possible values between adjacent units on a scale, it changes in fixed amounts (number of professors, cups of coffee)
-
-
Case Control
studies that begin after individuals have already developed or failed to develop the condition being investigated
-
Cohort
group of individuals who share a common experience
-
Randomized Clinical Trial (RCT)
- individuals assigned to study and control group
- "chance" (randomization) is used to assign to groups
-
Kurtosis
peakedness of a distribution
-
3 Types of Kurtosis
- 1. Mesokurtic (normal)
- 2. Leptokurtic (peaked)
- 3. Platykurtic (flattened)
-
Variance
- standard deviation squared
- -the higher, the more dispersed the data is around the mean
- SS/N
-
Standard Deviation
- how scores are dispersed around the mean
- square root of SS/N
-
Skewness
- symmetricality of a distribution
- can be positive or negative
-
Z Scores
- provides information about the location of scores in standard deviation units
- provides: direction and magnitude
- Mean= 0, SD=1
-
Central Limit Theorem
When the population distribution is not normal, the sampling distribution of the mean will be closer to the normal distribution than the population distribution
-
Comparing one group of subjects to a population
One Sample Hypothesis Testing
-
If a score falls within the Critical Region:
it is very unlikely to be just by chance
-
Random Error
- 95% of the population (in the middle of the distribution)
- -if Z score falls within -1.96 and +1.96, it is not statistically significant
-
Critical Regions
- 5% of population outside of Random Error
- 2.5% on each side
- -if Z score falls in critical region, it is statistically significant
-
Hypothesis Testing is
ruling out chance
-
Null Hypothesis assumes:
- that there will be no statistical significance
- Goal is to disprove/reject the Null
-
Type I Error
Saying something is significant when it's not (false positive)
-
Type II Error
Saying something isn't significant, when it is (fall negative)
-
One-Tailed vs. Two-Tailed
- One-Tailed: saying there is going to be a direction (this drug will increase attention)
- Two-Tailed: saying it's going to be different (there is a difference in IQ between men and women)
-
I Have A Crazy Teacher Dave stands for:
- Hypothesis
- Alpha Level (.05)
- Critical Region
- Test Statistic
- Decision
-
Independent/Between Measures t-Test
- Two independent group- one dependent group
- -Men vs. Women on time spent studying
- -Exercise/Don't exercise on how many hours they sleep
- -Study/Don't study on test scores
-
Levene's Test
- looks at if the variance between two groups is significantly different
- If it is below .05, they are significantly different (violation of your assumption)
-
A very difficult test will likely result in what type of skewed distribution?
negative
-
A sample's variability will be larger or smaller than the population from which it comes?
Smaller
-
The probability of Tye I Error is known as
Alpha
-
A Z Score has a mean of ____ and a standard deviation of _____.
- mean=0
- standard deviation=1
-
If raw scores from a positively skewed distribution were transformed into z-scores, what shape would the standardized scores assume?
Positive
-
When one analyzes data based on a sample, one calculates a _____________.
Statistic
-
Mathematical methods used to draw tentative conclusions about a
population based on sample data are referred to as _______ statistics.
Inferential
-
Which study can one most reasonably determine cause and effect?
Experiment
-
In Inferential Statistics, the object is to generalize from a _____________ to a ______________.
-
Which type of measurement: Sex of a Child
Nominal
-
Which type of measurement: Religion of an individual
Nominal
-
Which type of measurement: Time required to complete a task
Ratio
-
Which type of measurement: Ratings of a task (Easy, Mildly Difficult, Difficult)
Ordinal
-
In a normal curve, the mean will be ______________ the median and mode.
Equal to
-
In a normal distribution, approximately ________% of the scores fall
from one standard deviation below to one standard deviation above the
mean.
68%
-
The attitude of one group of employees differs from attitudes of the general population--what is the most appropriate hypothesis test?
One-Sample t-Test
-
If Levene's test is significant, the difference in means between the two groups _____________.
Cannot be determined
|
|
