Methods Chapter 2 Flashcards
Define Empiricism
accurate knowledge can be acquired trough observation
Empirical Method
a set of rules and techniques for observation
Scientific Approach
events governed by lawful order
Theory
hypothetical explanation a natural phenomenon
Hypothesis
Falsfiable PREDICTION made by a theory
Operational Definition
description of x in concrete, measurable terms i.e. “piercing - 1 piercing/person NOT 500 gross
Instrument
can detect CONDITION from Op. Definition
Validity
goodness of an event defining a property i.e. - 1 pierce has validity// ie frequency of smiling valid to determine happiness
reliability
instrument produces same measurement when measuring same thing i.e.. smile detected same way for TUES/THURS
Power
instrument’s ability to detect small magnitudes of a property
Demand Characteristics
aspect of setting/enivronment - causes people to behave as THEY THINK others WANT THEM to BEHAVE
Naturalistic Observation
unobtrusive - in natural environment - ie Jane Goodall chimps ie Hawthorne effect: ANY manipulation increase productivity - felt special
Examples of Participation Observation
Abu Biruni (973-1048 bad example) - FESTINGER “When Prophecy Fails” with Cognitive Distance Theory
These are the goals of the scientific approach
- Description-Measurement 2. Understanding-Prediction 3. Application-Control
Name the constructs of measurement
- self report measures - reports by others - behavioural observations - physiological measures
Define Variable
Variable - any MEASURABLE condition, event, characteristic, behaviour THAT IS manipulated, controlled, observed
Steps of Scientific Investigation
HSCAR 1 Hypothesis 2 Select Method 3 Conduct Study 4 Analyze Data-Draw Conclusions 5 Report findings of Research
Advantages of Scientific Method
Clarity/Precision -error free
The 2 research methods
1 Experimental - 2 Correlational
Define Experiment
Experiment - a technique to Establish CAUSAL RELATIONSHIP between 2 variables
Define Manipulation
Changing a variable in order to determine its casual power
Independent Variable
vairable Manipulated in an experiment
Experimental Group
exposed to the manipulation vs. control group not exposed - in example of speech manipulation TAPPING WAS CONTROL
Extraneous Variables (standardized variables)
others than independent that MAY Influence dependent variable
Confounding
value of extraneous variable varies across conditions of an experiment AKA changes along with the independent vble
Random Assignment
all subjects equal chance of being assigned to any Group/condition in the study
Internal Validity
attribute of experiment allowing to Establish CASUAL RELATIONSHIPS
External Validity
variables defined in typical way - simulates real world
Population (PSYC)
complete collection of participants who might be measured
Sample
partial collection of people drawn from a population
Case Method
a procedure to gather scientific information - studying single individual
Random Sampling
technique to ensure every member of population has chance to be included in the sample
define Between-subjects design
participant is in control or treatment group - NOT both - new group for each treatment
Multiple IV design
- either between- or within-subjects variables - e.g., effects of gender and drug on memory
Variations on experimental design
multiple dependent variables eg. examine both accuracy and speed of performance of data entry
Quasi-Experiment
any experimental design that DOES NOT control the influence of all extraneous variables - often arise in situations where it is not possible to randomly assign subjects - ie how gender effects locus of control need a specific gender
advantages of experimental method:
establish CAUSE/EFFECT relationship - control of extraneous variables
limitations of experimental method
- often artificial - lack of generalizability - ethical concerns - THREATS to INTERNAL validity
How are between subject designs kept random?
Randomization and matched pairs so very little contamination by extra variables but it’s still possible
Between subject designs disadvantages
Too many participants may be needed – groups still individuals – can skew – emotions can skew = confident
Example of between subjects experiment gone wrong
Researchers put school A as a control group School B as treatment - different socioeconomic zone each school is in causes bias
Explain within subject design experiments
Within-subject design - every subject is given every treatment including control ie placebo then Real drug later
Within-subject design advantages
- reduces amount of error from people variance - requires fewer participants
Within subjects design disadvantages
Fatigue - tired so performance effected in last test OR confident - better results on last test
explain multiple IV design
either between or within subjects variables eg effects of gender AND drugs on memory
some variations on experimental design
multiple dependent variables eg examine both accuracy and speed of performance of data entry
Quasi experiment
Any experimental design that does not control influence of all extraneous variables– Situations where it’s not possible to randomly assigned subjects i.e. gender affecting control
Advantages of experimental method
Establish cause-and-effect relationship - Control of extraneous variables
Limitations of experimental methods
o often artificial - lack generalizability - THREATS to internal validity still POSSIBLE
Types of correlation (3)
observation - case studies -surveys
Observational types
naturalistic, participant, blind, double blind
case study examples
Piaget - children develop ability to see split mass is same mass // Death row study - everybody had brain damage
problems With case studies
lack of generalizability
Survey problems
papaort and burkhart obsrved underreport of bad behaviour in survey participants
correlational method advantages
flexible - behaviour assessed in real world - allows for prediction of behaviour up
define matched samples
matched samples - two groups identical in terms of a third variable
define matched pairs
each participant is identical to one other participant in terms of a third variable
Correlational method limitation
Cannot establish causality
Explain bidirectional problem
birthrate causing religion OR religion causing birthrate? could be either way
Explain third variable problem
PArenting is influencing both birth rate and religiosity
Example of sampling bias
literary digest predicts Landon over Roosevelt - phone users sampled + rich + republican
Placebo effect
- have placebo control group so its not causing it
THIS decreases internal validity
confounding due to extra variables
List distortions of self report data and experimenter bias
self report - subjects skew to SOCIAL DESIRABILITY Experimenter bias - double blind procedure eliminates observers expectations - rosenthal
these were ethics codes set up
nuremburg codes - after NAzis did torture experiments
This was an “evil” US study
Tuskegee experiment - Syphilis patients - began 1932 to 1972!!! US Public Health Service refuses black patients treatments and observed
define debriefing
verbal description of purpose of study AFTERWARDS
define informed consent
written agreement to particicpate in the study
describe APA code of ethics
1953: competence, integrity, professional academic responsibility, respect for rates, concern for others welfare social responsibility
Describe CPA code of ethics
Respect for dignity of persons, responsible caring, integrity in relationships, responsible to society
Canada tri council policy statement
Respect for human dignity, free consent, vulnerable persons, confidentiality, justice inclusiveness, BALANCING HARM & BENEFIT
outline steps to design ethical study
Obtain informed consent - allow subjects to withdraw, data confidentiality, debrief after - desensitize
Animal research examples
Sperry - split brain right/left - Biofeedback - Schanberg - newborn rats touching causes growth
Points against animal research and APA animal research code
-specieism fundamental rights, APA code caring for animals - justifying study
Double blind
Observation true purpose Hidden from the observer and subjects
Descriptive statistics
Summarize some aspect of the data
Frequency distribution
Graphical representation of measurement - # of times each measurement made = arrangement
Histogram
rectangles in a frequency distribution
explain Stem + leaf plot
ie 7 . 17 reads 71 & 77
Explain the three measures of central tendency
1 - mean - average 2 mode - most frequent score 3 Median midpoint ie 3,4,5 = 4, 3-4 = 3.5
Keep this in mind when choosing a measure of central tendency
MEAN affected by extreme scores - not others
Explain three measures of variability
1 range highest to lowest 2 Variance - average of squared deviations from mean i.e. 1^2 + 3^2 = 10 3 standard deviations sqrt(variance - good units
correlation
Wind variations in value or variable one synchronize with variable 2 variations
correlation coefficient
measure of DIRECTION and STRENGTH of correlation - letter r -1 to +1
explain various meanings of r
Perfect positive = +1 / perfect negative = -1 / none = no systemic change
explain normal distribution
in middle mean + median+mode - 68% scores within 1 standard deviation - 95% within 2 SDs
inferential statistics
GENERALIZATION from findings in a sample population - limit on errors - has significance
review parts to good experiment
1 Manipulative IV 2 randomly assign subjects 3 Measure Dependent V
Internally valid vs external
INTERNAL - casual relationship established - most PSYC not externally valid
limits of correlation
cannot confirm related - need controlled experiment