-
Where do beliefs come from?
- Tenacity
- Authority
- Experience
- Empirical Evidence
-
Empirical Evidence
- Systematic or formal observaion to obtain objective, reliable, valid and quantitative measures of the matter of interest.
- By itself, empiricism CANNOT explain WHY
-
Syllogisms
Example: All As are Bs. C is an A, therefore C is a B.
A syllogism is only valid if it obeys the rules of logic. Valid logic enables very strong truth claims GIVEN the empirical validity of the premises.
-
Common logical reasoning errors
- belief bias
- conversion errors
- confirming evidence bias
-
Science combines Rationalism and Empiricism. Define.
Rationalism: used to develop theories and hypotheses and way to test.
Empiricism: the means of conducting the tests.
-
What are the 4 features of Scientific Method?
- 1. Objectivity
- 2. Replication
- 3. Self-Correction
- 4. Control
-
Define Control as related to the Scientific Method
- Two meanings:
- 1. Directly manipulating the variable of interest
- 2. Controlling for unwanted variables that could influence the results.
This is essential to draw conclusions about cause and effect.
-
List and define the two "other" variables.
Subject Variables: individual differences in participants such as age, gender, IQ, ethnicity.
Quasi-independent variables: variables outside of the participant that the researcher cannot manipulate, such as weather, laws, geographical location.
-
What is an Extraneous Variable?
Other unwanted, uncontrolled factors that could influence the dependent variable. Confounds which invalidate the experiment.
-
Statistics
- Set of procedures for reducing large masses of data to manageable proportions in order to draw conclusions from those data.
- Two types: Descriptive and Inferential.
-
Descriptive Statistics
Inferential Statistics
Descriptive: Numbers that summarize a set of data.
Inferential: Calculations that determine whether an IV has a significant effect; they allow us to draw inferences.
-
Population
The complete set of events being studied. It is the entire group to which we want to generalize the results.
-
Parameters
Numerical values summarizing population data.
-
Sample
The subgroup of the population that we collect data from.
-
Random Sample / Random Selection
A sample in which each member of the population has an equal chance of inclusion in the study.
-
Convenience Sample
Participants selected for their accessibility or ease of testing.
-
Random Assignment
Everyone in the study has an equal chance of being assigned to each of the study groups.
-
What are the 2 types of data?
Continuous data (aka "measurement" or "quantitative" data). A mean can be determined.
Categorical (aka "frequency" or "count" data). No mean is possible.
-
List the 4 scales of measurement.
- Nominal
- Ordinal
- Interval
- Ratio
(they go from simple to complex)
-
Notation
N = ?
n = ?
X = ?
- N = total sample size
- n = number of participants per group
X = set of scores for one variable
-
Four principles from most to least important of the CPA
- Respect for the Dignity of Persons
- Responsible Caring (competence)
- Integrity in Relationships (honesty)
- Responsibility to Society
-
Respect for Dignity of Persons entails...
- Privacy
- Confidentiality
- Informed Consent
- No harassment or degrading comment
- No unjust discrimination
- Fair compensation
-
Responsible Caring entails...
- Protect welfare of others, avoid harm
- Take responsibility for actions
- Keep up to date
- Referral for help
- Maintain appropriate relationships
- Pilot studies
-
Integrity in Relationships entails...
- No dishonest, fraud or misrepresentation in reporting results
- Do not supress disconfirming evidence
- Acknowledge limitations of findings
- Do not deceive if not necessary
- Debriefing
- No coercive enticement to participate
-
Responsibility to Society entails...
- Contribute to discipline and state of knowledge
- Keep informed
- Critical self-evaluation
- Educate & promote scientific growth of others
- Respect for social customs, cultural expectations
- Sensitive to needs of society when designing research (hot topics)
-
Tri-Council Ethical principles (for all disciplines)
- Respect for Human Dignity (cardinal principle)
- Respect for Free and Informed Consent
- Respect for Vulnerable Persons
- Respect for Privacy and Confidentiality
- Respect for Justice and Inclusiveness
- Balance of Harms and Benefits
-
Explain Debriefing
Explanation of purpose of study, methods used, correct misconceptions, ask if any questions
- Education: What was the study about?
- Dehoaxing: Describe deceptions and why?
- Desensitizing: Any psychological discomfort?
-
Why is it important to survey the literature?
- To determine the current state of the knowledge
- Provide a basis for hypotheses
- Guide you in selecting paradigm, operational definitions
-
List the steps to beginning research.
- Step 1: Develop a research question
- Step 2: Survey the literature
- Step 3: Build a hypothesis
-
What are the characteristics of the research hypothesis?
- Synthetic statement: is either true or false
- Falsifiable: can be shown to be wrong
- Can be stated in "General Implication Form" (if...then...)
- Can be directional (more/less than) or non-directional (different from)
-
Inductive vs. Deductive logic
Deductive: General to specific; How we form our research hypotheses
Inductive: Specific to general; Combining the results of several studies into a theory
-
List 4 types of independent variables
- physiological (manipulation of biological state)
- experience (manipulation of amount/type of training/learning)
- stimulus/environmental (manipulation of the environment)
- participant (manipulation of aspects of participant)
-
List the types of dependent variables
- correctness
- rate/frequency
- degree or amount
- latency or duration
-
Nuisance variables
- Unwanted variables that increase the variability of all scores within groups
- Affects ALL groups
- Makes the effect harder to see
-
Confounders (aka Extraneous variables)
- Unintended influences on the DV
- biases result in a particular direction
- renders findings MEANINGLESS
-
Developing good controls for extraneous variables
- Step 1: Randomization
- Step 2: Elimination (of extraneous variables)
- Step 3: Constancy (across all groups of participants)
- Step 4: Balancing (equal distribution of extraneous variables)
-
What is an order effect?
- When the position in a series affects how participants respond.
- Doesn't depend on the EVENT but on the POSITION
- think fatigue/practice/learning
-
What is the carryover effect?
- When the effects of one event influences responses to the next event.
- Depends on the EVENT not the POSITION
- (i.e., previous drug intake)
-
Complete Counterbalancing
works to counteract carryover and order effects:
- 1: Each event must be presented to each participant an equal number of times.
- 2: Each event must occur an equal number of times at each session.
- 3: Each event must precede and follow each of the other events an equal number of times.
-
Every measure consists of two elements:
- True score (hypothetical concept); and
- Error (bias and random error)
-
Observed score =
Observed score = True Score + Error (bias + random)
-
Experimentor error may be...
- Random error: noise, temp., time of day.
- Bias error: experimenter characteristics or experimenter expectancies (Rosenthal)
-
How do we control for experimenter characteristics?
- Use standardized methods
- Replication
-
How do we control for experimenter expectancies?
- Standardization
- Objectivity
- Single-blind research
-
Participant Error
Random participant error: carelessness, distraction
Participant Bias: Demand characteristics, Good Participant effect, response bias
-
Define demand characteristics
Features of an experiment that seem to inadvertently cause participants to act in a particular way.
-
Define Good Participant Effect
Tendency for participants to behave as they think the researcher wants them to behave.
-
How to control for Demand Characteristics?
- Condcut double-blind research
- Use deception
-
What is response bias?
- Yea- and Nay-sayers.
- When the context affects participant response
- Can be a factor of the experimental setting or the questions
- Social desirability can be an issue
-
How to control for Response Bias?
- Include "agree" and "disagree" items
- Randomize question presentation.
- Pilot testing
-
Describe Observer Error.
- Random observer error: carelessness, distraction
- Observer/scorer bias: confirmatory bias
More important to reduce observer bias (confound) than random error (nuisance).
-
How to control for Observer Error?
- Eliminate human observer (use mechanical measure to reduce random and bias errors)
- Limit observer subjectivity (focus on observable behavior, standardized coding)
- Make observer blind
-
What is construct validity and list 4 components.
- Does the manipulation or measure ACTUALLY represent the claimed construct?
- Reliability
- Content validity
- Convergent validity
- Discriminant or divergent validity
-
How to establish Reliability?
- assess random error
- reliability is a prerequisite for validity
- Test-retest reliability
- Inter-rater reliability
- Internal consistency
-
Describe Internal Consistency
- measure of participant random error
- variability across items = random error
- Index calculated using Cronbach's Alpha, split-half correlation and average inter-item correlation
-
What is content validity?
- Is the measure's content relevant to the concept?
- Does it clearly relate to the concept?
- Does it cover all aspects of the concept?
-
What is convergent validity?
Does measure correlate with other indicators of the same construct?
-
What is Discriminant Validity?
Is the measure distinguishable from other constructs?
-
What is Sensitivity?
- Sensitivity is the ability of measures to detect effects.
- "does your measure minimize the influence of error?"
- Use measures with maximal validity and maximal reliability
- Avoid restriction of range, all or nothing measures.
- Add scale points to a rating scale
- Pilot test measure
-
Why conduct non-manipulation studies?
- Naturlistic research settings;
- manipulation not possible
- natural variation
- prediction and selection
- temporal change
- comparing size of associations
-
List the types of descriptive studies
- Archival research
- Observational techniques such as case studies,
- naturalistic observation,
- participant observation
- clinical perspective
-
What are some issues with Archival Studies?
- Limits generalization
- may have missing data values
- may not be ideal to your research question
- cannot show causation
-
What are the differences between Clinical Perspective and Participant Observation?
- client chooses clinician, whereas participant observer chooses others to study
- clinicians cannot be unobtrusive or passive
- Participant observer's goal is understanding whereas clinician's goal is helping
-
What are some issues with Observational Techniques?
- Reactivity: when the knowledge of being watched affects behavior, aka "The Hawthorne Effect"
- High on external validity but low on internal validity
- Cannot make cause-effect statements
- objectivity
-
What are descriptive surveys?
Descriptive surveys seek to determine what % of the population have particular characteristics, beliefs or behaviors.
-
What are analytic surveys?
Analytic surveys seek to determine the relevant variables and how they are related.
-
What is Cronbach's Coefficient Alpha?
- Most common estimate of test reliability
- measures how well a group of items measure one uni-dimensional construct
- should be at least 0.7
-
What is the difference between tests/inventories and surveys/questionnaires?
- surveys/questionnaires: examine an opinion
- tests/inventories: assess a specific attribute, characteristic or ability of the subject
-
A good test should have VALIDITY which is established by:
- content validity
- concurrent validity
- criterion validity: can the test predict future behavior?
-
A good test should have RELIABILITY which is established by:
- test-retest consistency
- split half consistency
-
Samples can be collected by:
- Random sampling; or
- Stratified random sampling: the population is separated into different subgroups and a random sample is taken from each
-
What are 3 research strategies?
- Single-strata: select from a subgroup of the population
- Cross-sectional: multiple subgroups at the same time
- Longitudinal: one cohort over an extended period of time
-
What is qualitative research?
- An attempt to capture the complexity of human behaviour in its natural environment.
- INDUCTIVE
-
What is Positivism?
philosophical position that stresses observable facts and seeks universal laws
-
What is "post-positivism"?
both quant and qual research concerned with the collection of observed information but NOT universal laws.
-
What is Grounded Theory?
Attempts to use qualitative methods to identify themes and build a theory.
-
What is correlational research?
- both a statistical technique and a research method
- research designed to determine whether an association exists between 2 variables
|
|