experiment 0 black screen slides

Post on 12-Jan-2016

27 Views

Category:

Documents

0 Downloads

Preview:

Click to see full reader

DESCRIPTION

Experiment 0 Black screen slides. Tiger Apple Television Balloon Night Water Orange Jacket Soda Radio Soap Lamp. List 1 – No delay. Tiger Apple Television Balloon Night Water Orange Jacket Soda Radio Soap Lamp. List 2 – 30 sec delay. Soap Opera New York Academy - PowerPoint PPT Presentation

TRANSCRIPT

Experiment 0Experiment 0

Black screen slidesBlack screen slides

List 1 – No delay List 1 – No delay

• TigerTiger• AppleApple• TelevisionTelevision• BalloonBalloon• NightNight• WaterWater• OrangeOrange• JacketJacket• SodaSoda• RadioRadio• SoapSoap• LampLamp

List 2 – 30 sec delayList 2 – 30 sec delay

• TigerTiger• AppleApple• TelevisionTelevision• BalloonBalloon• NightNight• WaterWater• OrangeOrange• JacketJacket• SodaSoda• RadioRadio• SoapSoap• LampLamp

List 3 – 30 sec delayList 3 – 30 sec delay

• SoapSoap• OperaOpera• NewNew• YorkYork• AcademyAcademy• AwardAward• EastEast• WestWest• NorthNorth• SouthSouth• MorningMorning• NightNight

List 4 – No delayList 4 – No delay

• ThatThat• WithWith• FarFar• ThenThen• ThroughThrough• AboutAbout• WhichWhich• SinceSince• ByBy• UntilUntil• AmongAmong• UponUpon

RESEARCH VARIABLESIndependent variable (IV): Variable manipulated by

researcher

Dependent variable (DV): Variable measured by researcher, expected to change as a result of researcher’s manipulation (“depends on IV”)

Control variables (CV): Factors kept constant (e.g., confederate, room).

Random variables (RV): Factors allowed to vary randomly.

(e.g. time of day effects, fatigue, motivation)

Confounding variable: Variable’s levels covary with IV’s and may be responsible for changes

Research in PsychologyResearch in Psychology

WHAT DRIVES PSYCHOLOGICAL RESEARCH?WHAT DRIVES PSYCHOLOGICAL RESEARCH?- Existing Theories: Evaluate, extend, or discreditExisting Theories: Evaluate, extend, or discredit- Other Research: Replicate, extend, or discredit Other Research: Replicate, extend, or discredit

- Personal Observation Casual curiosity or new - Personal Observation Casual curiosity or new interest interest

Scientific MethodScientific Method

Evaluate ideas objectively, with degree of Evaluate ideas objectively, with degree of confidenceconfidence

Rules on how to collect, summarize, and analyze Rules on how to collect, summarize, and analyze data data

Evaluating Without Scientific Evaluating Without Scientific MethodMethod

““Clever Hans” phenomenon, facilitated autistic Clever Hans” phenomenon, facilitated autistic communicationcommunicationNeither knew answer if questioner failed to know itNeither knew answer if questioner failed to know it

Hans was sensitive to primate nonverbal cuesHans was sensitive to primate nonverbal cues

Science is a set of rules to Science is a set of rules to keep us from lying to each keep us from lying to each

otherother

Experiments allow conclusion about causality:Experiments allow conclusion about causality:

ControlControl: Keep variables & groups constant or : Keep variables & groups constant or equivalentequivalent

ManipulationManipulation: Change only one thing at a time: Change only one thing at a time

• Scientific progress is achieved Scientific progress is achieved incrementally.incrementally.– Share findings with scientists and public via Share findings with scientists and public via

peer-reviewedpeer-reviewed journals, or conferences, or journals, or conferences, or books.books.• Peer reviewPeer review takes time (up to 1 year after takes time (up to 1 year after

submission), but allows other experts to evaluate and submission), but allows other experts to evaluate and critique your work, occasionally revealing flaws or critique your work, occasionally revealing flaws or limitations of the study.limitations of the study.

• HOW TO FIND JOURNAL PAPERS – PubMed & PsycInfoHOW TO FIND JOURNAL PAPERS – PubMed & PsycInfo

Low to High ConstraintLow to High Constraint

• Correlational designCorrelational design– ObservationalObservational– Case StudyCase Study– SurveySurvey– Archival researchArchival research

• Experimental designExperimental design– Quasi-ExperimentQuasi-Experiment– ExperimentExperiment

• QUANTIFICATION OF BEHAVIOR QUANTIFICATION OF BEHAVIOR

• Descriptive Statistics:Descriptive Statistics: – Characteristics of your sampleCharacteristics of your sample

• Central tendencyCentral tendency: : – Mean, Median, ModeMean, Median, Mode

MeanMean

• Arithmetic averageArithmetic average

• 11• 11• 66• 55• 22• Mean = 15 / 5 items = Mean = 15 / 5 items = 33

• MeanMean

• Kirkwell WA per capita income: Kirkwell WA per capita income: $1,200K$1,200K

• Mean = sensitive to extremesMean = sensitive to extremes

MedianMedian

• Arrange in order, middle value is medianArrange in order, middle value is median

• 11• 11• 2 2 • 55• 66• Median = Median = 22

Median when even # of Median when even # of itemsitems• Arrange in order, middle value is medianArrange in order, middle value is median• (if even number of items, average two middle values)(if even number of items, average two middle values)

• 11• 11• 2 2 • 5 5 • 66• 77• Median = (Median = (2+5)/2 = 3.52+5)/2 = 3.5

MedianMedian

• Not sensitive to Not sensitive to extremesextremes

• 11• 11• 2 2 • 55• 66• Median = Median = 22

• 11

• 11

• 2 2 • 55

• 6 billion6 billion

• Median = Median = 22

• -150-150• 11• 2 2 • 55• 66• Median = Median = 22

• MedianMedian

• Kirkwell WA per capita income:Kirkwell WA per capita income:$55K$55K

Median is nonlinear, Median is nonlinear, affected by groupingaffected by grouping

• 11

• 11

• 44

• Median =1Median =1

• 44

• 44

• 1616

• Median =4Median =4Median for all numbers = (1+4)/2 = 2.5

• 11

• 11

• 44

• 44

• 44

• 1616

• Median =4Median =4

ModeMode

• Most frequent value Most frequent value

• 1 1 • 1 1 • 22• 55• 66• Mode = 1Mode = 1

• ModeMode

• Kirkwell WA per capita income: Kirkwell WA per capita income: $13K$13K

Mode is also nonlinear, Mode is also nonlinear, affected by groupingaffected by grouping

• 11

• 11

• 44

• Mode =1Mode =1

• 44

• 44

• 1616

• Mode =4Mode =4Mode for all numbers = (1+4)/2 = 2.5

• 11

• 11

• 44

• 44

• 44

• 1616

• Mode =4Mode =4

How well central tendency describes How well central tendency describes data? - need index of dispersiondata? - need index of dispersion

• 11

• 22

• 33

• 44

• 55

• 66

• 77

• 88

• 99

• Mean = 5Mean = 5

• 44

• 44

• 55

• 55

• 55

• 55

• 55

• 66

• 66

• Mean = 5Mean = 5

DispersionDispersion:: Range, Standard DeviationRange, Standard Deviation

• 11

• 22

• 33

• 44

• 55

• 66

• 77

• 88

• 99

• Mean = 5, Range = 8Mean = 5, Range = 8

• 44

• 44

• 55

• 55

• 55

• 55

• 55

• 66

• 66

• Mean = 5, Range = 2Mean = 5, Range = 2

DispersionDispersion:: Range, Standard DeviationRange, Standard Deviation

• 44• 44• 44• 44• 55• 66• 66• 66• 66• Mean = 5, Range = 2Mean = 5, Range = 2• St Dev = 0.94St Dev = 0.94

• 44• 44• 55• 55• 55• 55• 55• 66• 66• Mean = 5, Range = 2Mean = 5, Range = 2• Std Dev = 0.67Std Dev = 0.67

Measurement ScalesNominal or Nominal or

categoricalcategorical OrdinalOrdinal IntervalInterval RatioRatio

People with the People with the same scale same scale value are the value are the same on some same on some attribute. attribute.

Scale values have Scale values have no 'numeric' no 'numeric' meaning in the meaning in the way that you way that you usually think usually think about numbers.about numbers.

People with a People with a higher scale higher scale value have value have more of some more of some attribute. attribute.

Intervals between Intervals between adjacent scale adjacent scale values are values are indeterminate. indeterminate.

Intervals between Intervals between adjacent scale adjacent scale values are values are equal with equal with respect to respect to attribute being attribute being measured. measured.

E.g., difference E.g., difference between 8 and between 8 and 9 is the same as 9 is the same as the difference the difference between 76 and between 76 and 77.77.

Meaningful zero Meaningful zero point for the point for the scale. scale.

Ratios are Ratios are equivalent, e.g., equivalent, e.g., the ratio of 2 to the ratio of 2 to 1 is the same as 1 is the same as the ratio of 8 to the ratio of 8 to 4.4.

NominalNominal OrdinalOrdinal IntervalInterval RatioRatio

Gender. Gender. Ethnicity.Ethnicity.Marital Status.Marital Status.  

SESSESUSDA beef ratings USDA beef ratings (good, choice, prime)(good, choice, prime)HospitalizationHospitalizationTop 20 CountdownTop 20 Countdown

Degrees F. Degrees F. Letter Grades Letter Grades IQIQ  

Degrees K. Degrees K. Annual income Annual income TimeTimeDistanceDistance

Role of FiguresRole of Figuresin Scientific Papersin Scientific Papers

• Argument made in pictures, not Argument made in pictures, not words!words!

Figures used for:Figures used for:

• DATADATA

• METHODMETHOD

• THEORYTHEORY

Three mountain test for Three mountain test for egocentrismegocentrism

Hierarchical memory Hierarchical memory storage storage

Lies, damn lies, and Lies, damn lies, and statisticsstatistics

eyetrackingeyetracking

Ages, wealth, length of Ages, wealth, length of absenceabsence

RESEARCH VARIABLESIndependent variable (IV): Variable manipulated by

researcher

Dependent variable (DV): Variable measured by researcher, expected to change as a result of researcher’s manipulation (“depends on IV”)

Control variables (CV): Factors kept constant (e.g., confederate, room).

Random variables (RV): Factors allowed to vary randomly.

(e.g. time of day effects, fatigue, motivation)

Confounding variable: Variable’s levels covary with IV’s and may be responsible for changes

DESIGN

Within-subject design - each participant exposed to every level of IV (person acts as own control)

Between-subject design - each participants exposed to one level of IV (and unaware of what other participants are exposed to)

Counterbalancing - ordering presentation levels of IV to eliminate confounding

OPERATIONAL DEFINITION

Specify operations required to manipulate or measure concept.

e.g. Anxiety is not vaguely measured by objectively through scales, checkboxes, etc

VHS 1518.02 Final pointsVHS 1518.02 Final points

• Correlation is not causationCorrelation is not causation– Ice cream and murders in NYCIce cream and murders in NYC

• Single-blind and double-blind designsSingle-blind and double-blind designs– Knowledge of which level of the IV subject is Knowledge of which level of the IV subject is

assigned toassigned to

• Controlled experimentControlled experiment– Active treatment vs non-active “treatment” Active treatment vs non-active “treatment”

such as a placebosuch as a placebo

top related