Research Methods Flashcards
Why isn’t true score = score?
Random vs systematic measurement error
Score (X) = T (True score) +/- E (Measurement error)
- Measurement error effects
—
Random: Random fluctuations, influences in unpredictable ways, unrelated to construct
Systematic: Non-random fluctuations, influences in predictable ways, related to construct
Can a test be reliable but not valid?
No
Must be reliable to be valid
Reliability
- Test-retest (temporal stability, retest correlation)
- Inter-rater (kappa)
- Internal consistency (cronbach’s alpha)
Validity
- Content, construct, criterion, internal/external validity
Replicability
Consistent information; Tests random measurement error
- Long period w/ high correlation good, 0.7+ correlation between two scores
- High rater agreeability, 0.7+ good
- Test items correlated to each other, 0.7+ good
—
Measures what it’s supposed to measure (tests systematic measurement error, true variance)
- Content measures what it’s supposed to; Operationalization done properly and matches factor analysis; Content correlates w/ outcome; How certain DV is changing from IV; Results can be generalized to real world
—
Extent to which results can be found again in context diff from og researxy
Classification
Talents vs Strengths (Cliffton Strengths Finder)
- Validity/Accuracy + Pros, Limits
Way to name, organize, categorize diversity of behavs/traits
—
- Talents: Naturally occurring patterns of thought/feelings/behav that can be productively applied + manifested in life exps characterized by desires; trait-like properties resulting from successful development
- Strengths: Extension of talent, combines w:/ assoc knowledge and skills
- Test-retest, internal consistency, construct validity, discriminant validity + has high utility in diff ages/enviros BUT not sensitive to true changes in talents
Values in action (VIA)
- Character strengths (How it differs from Strength Finder)
- Strengths (6) + 2 virtues each
- Character and signature strength criteria (2 each)
- psychometric properties + pros, limitations
Character strengths: Psychological processes/mechanisms that define virtues on a continuum
—
Wisdom/Knowledge: Cog strength to acquire and use knowledge
- Creativity, curiosity
Courage: Emotional strength to accomplish goals in face of opposition
- Bravery, persistence
Humanity: Interpersonal strength, tending/befriending others
- Kindness, love
Justice: Civic strengtht underlying healthy community life
- Fairness, leadership
Temperance: Protects against excess
- Forgiveness, modesty
Transcendence: Forges connections to larger universe + meaning
- Gratitude, religiousness
—
- Character strength: Trait-like, be absent from some ppl
- Signature strengths: Sense of ownership, positive feeling when using
- Test-retest reliability, convergent validity, discriminant validity + sensitive to change BUT lacks construct validity (factor analysis shows diff virtues)
Search Institute’s 40 developmental assets help adolescents thrive by what external and internal assets?
External: Positive exps gained thru interactions w/ ppl and institutions
Internal: Personal characteristics/behavs that stimulate positive devel
Hedonic vs Eudaimonic well-being
Psychometric analysis of:
- Positive and Negative Affect Schedule (PANAS)
- Satisfaction with Life Scale (SWLS)
Hedonic: Experiencing high lvls of pleasure and low lvls of displeasure
- Subjective well-being
Eudaimonic: Actualizing one’s potentials or finding a sense of purpose in life
- Psychological well-being
—
PANAS: Good reliability, convergent + discriminant validity BUT unsure if measuring moods or emotions
SWLS: Good internal consistency, test-retest, consistent across organizations + measures true changes
6 components of Ryff’s Psychological Well-being Scale
(LGRAMA)
Purpose in life: Have goals + sense of direction
Personal growth: Feelings of continued devel and potential
Positive relations w/ others: Have trusting relationships and concern about others’ welfare
Self-acceptance: Positive attitude toward self
Environmental mastery: Feel competent and able to manage enviro
Autonomy: Understanding/Viewing self ad self-determining
4 components of Keyes’ Social Well-being scale (Related to social challenges)
(AACI)
Social acceptance: Positive attitudes/acceptance towards ppl
Social actualization: Care about/believe society is evolving positively
Social contribution: Feel one had value to give to society
Social integration: Feel part of community
Seligman’s PERMA model
Keyes’ Complete State Model of Mental Health
Positive emotions: High lvls of positive affect
Engagement: Activities leading to flow state
Relationships: Positive relationships
Meaning: Pursuing meaningful expa
Accomplishment: Achieving goals
—
Low to high well being, high to low mental disorder
- Flourishing: Low disorder, high well being
- Struggling: High disorder, low well being
- Floundering: High disorder, high well-being
- Languishing: Low disorder, low, well-being
Research methods:
- Archival research
- Case study
- Qualitative research
- Ethnography
- Correlational research
Re-examining existing records for new purpose
- Adv: Study changes/Contexts in time, less expensive
- Dis: Only correlational, can’t control accuracy/validity
—
Record of individual’s exps/behavs/both by an outside observer
- Adv: Detailed description, can study rare cases, provides exceptions to accepted ideas, good training tool
- Dis: Only correlational, low external validity, low reliability
—
Record of small group’s exps/behavs kept by an outside observer
- Adv: Detailed description, illuminates important issues/themes
- Dis: Only correlational, Low validity, low reliability
—
Understand unique values/traditions/social processes of culture by living w/ its members + making external observations
- Adv: Detailed descriptions
- Dis: Internal validity varies, low external validity
—
Looks at links between two variables in the world
- Adv: Looks at links of events in real world
- Dis: Only correlational, directionality problems, third variable problems
Correlational designs:
- Cross-sectional
- Longitudinal
(Research methods)
Assessment of several groups of ppl at same point in time
- Adv: Can find age diffs, economical, less time consuming, large samples
- Dis: Can’t tell age changes, cohort effects (diffs might be caused by diffs in time born)
—
Assessment of same group of ppl at regular intervals over time
- Adv: Can find age-changes and intra-individual stability, causal inferences allowed, no cohort effects
- Dis: Costly, time consuming, attrition high, time effects/cultural shifts can happen, third variable problem
Extperimental research
Quasi-experimental research
(Research methods)
Influence of IV on DV thru random assignment and control of extraneous variables
- Adv: Causal inferences
- Dis: Low external validity (mundane/psychological realism), some variables may be difficult to change
—
Look at diffs between ppl in naturally-occurring groups
- Adv: High external validity, good when variables unethical to manipulate
- Dis: Low internal valudity
