-
Definition of Statistics
-
Set of mathematical procedures for research in the behavioral sciences
- Organize & summarize information
- Determine what conclusions are justified
- Parameter: Characteristic or value that describes a population
- Statistic: Value or characteristic from a sample
-
Populations & Samples
- Population: Everyone of interest in a particular study
-
Sample: Set of individuals selected from a population
- Researcher wants to generalize back to the population
-
Variable: characteristics or condition that can have different values (change)
- Must be able to measure
-
Datum or score or raw score: individual measurement or observation
- Data set: collection of data
- Statistics is about relationship between sample statistics & population parameters
-
Data Structures, Research Methods, & Statistics
- Correlational Method
- Experimental method
- Non-experimental method
-
Variables & Measurement
-
Constructs
- Internal attributes that cannot be directly observed
-
Operational Definitions
- Describes set of operations for measuring a construct
- Defines the construct in terms of the resulting measurements
-
Discrete Variables
- Separate, indivisible categories (whole, countable numbers)
-
Continuous Variables
- Can be divided into infinite number of fractional values
-
Real Limits
-
Upper real limit
- Top of the interval
-
Lower real limit
- Bottom of the interval
- Boundaries of intervals for scores on a continuous number line
-
Scales of Measurement
-
Nominal Scale
-
Classification into categories
- No direction or size to difference (cannot quantify)
- Majors in school, gender, race
-
Ordinal Scale
-
Categories in a fixed order with size or magnitude differences but cannot measure/quantify difference
- Small, medium, large; rank orders, likability, preference
-
Interval Scale
-
Categories where intervals are exactly the same size, can measure/quantify difference & direction
- but zero is arbitrary, like temperature
-
Ratio Scale
-
Categories where intervals are exactly the same size, can measure/quantify difference & direction
- zero has meaning, represents absence of variable
- Ex. number of errors, time to respond
-
Statistical Notation
-
Summation notation
- Greek letter sigma
-
Descriptive & Inferential Statistical Methods
-
Descriptive statistics: Statistical procedures used to simplify & summarize data
- Summarize data
- Organize data
- Simplify data
-
Inferential Statistical methods: Used to study samples & make generalizations about populations
- Sampling error
- Margin of error
-
Experimental Method
- Purpose: Demonstrate cause & effect relationship between two variables
-
Experimenter manipulates one variable to see its effect on another variable
-
Must control for differences if want to use to generalize
- Characteristics of participants
- Characteristics of environment
- Control through: random selection of participants
- Control through: random assignment of participants to treatments
- Matching of participants
- Hold some variables constant
- Second (dependent) variable is measured or observed
- First variable (independent) is manipulated
- Control Condition: Control group which receives no treatment or placebo treatment
-
Correlational Research Method
- Both variables are measured to see if they have a relationship