Hypothesis Development
Forming a Hypothesis
A hypothesis is a statement predicting the results of a study (best educated guess)
Can be descriptive or casual
Often based on theories and results already known
Using the Literature to Form Hypotheses
Descriptive hypotheses are predictions about the results of a study that describe the behavior or the relationship between behaviors
Casual hypotheses are predictions about the results of a study that includes the causes of a behavior
Must be precise and logical
Must be testable
Reliability = consistency
Using the same procedure to measure the same phenomenon and obtaining the same results
A reliable measure is consistent between two or more separate measurements
Must be clearly specified!
Validity
Internal validity is the degree to which a study provides casual information
The design of experiment ensures the IV (independent variable) caused the measured difference in the DV (dependent variable)
“Was the research done right?”
External validity is the degree to whch results can be generalized beyond the specifics of your sample
“Does this happen in other settings? Everyday, labs?”
Data Collection Techniques
Naturalistic observation includes unobtrusively observing individuals in their natural/normal environment
Important to develop a coding scheme (guideline on how to pull relevant information)
Ethograms are lists of defined behaviors
Inter-observer/inter-rater reliability
Measure of the degree of agreement between 2 independent observers
Surveys/Questionnaires
Asks individuals to self-report their behaviors and attitudes
Questions need to be tested for reliability and validity
Likert Response Scale
(1 = Strongly Disagree → 5 = strongly agree)
Interviews
Requires coding scheme
More flexible w/questions
Systematic Observations
Observing behavior under controlled conditions
Includes memory accuracy, problem solving ,speed, and Implicit Association Test
Archival Data
Observing behavior from pre-existing records/data
Content analysis
Observation technique that invlves analysis of what someone has said or written
Research Designs
The specific method a researcher uses to collect, analyze, and interpret data
3 Main Types
Descriptive research (case study)
Correlational research
Experimental research
Descriptive Research
Descriptive studies seek to answer…
Who, what, where, when, how
When are these studies appropriate?
Understand prevalence and trends
Explor a phenomenon in depth
Examining a phenomenon in a different population
Case Study
A research design that involves intensive study of an individual or group of individuals and their behaviors
Often use naturalistic observations, interviews, or archival data to learn about behavior
Good example is Henry Molaison
Correlational Research
Examines the relationships between multiple dependent variables w/o manipulation of variables
Main goals include…
Determine if different behaviors are connected and occur or vary together
Predict a change in one from observing change in another
2 Types of relationships
Positive correlation (0 to +1.00)
A direct relationship, two things increase together or decrease together
Negative correlation (0 to -1.00)
An inverse relationship (they do the opposite) as one thing increases, the other decreases
Problems with Causality
Correlational research cannot be used to draw conclusions about the casual relationships among variables
Third variable problem: Presence of extraneous factors in a study that affects the dependent variable, decreasing internal validity
CORRELATION DOES NOT EQUAL CAUSATION
Experimental research
Assesses the impact of an experimental manipulation (independent variable) on an outcome of interest (dependent variable)
Control for confounding variables by using random assignmet, so you can assess cause and effect
Independent Variable
Manipulated by the researcher (can also be held constant or added to)
Levels of IV: Different conditions participants experience
Dependent Variable
The variable being measured/observed
Control group
Group of participants that do NOT receive the experimental treatment/manipulation
Quasi-Experiment Research
A comparison is made but participants are NOT randomly assigned to conditions
Not all variables can be manipulated
Use naturally occurring groups