Scribd Logo
Upload

Welcome to Scribd!

  • 44 views

    5 Alternatives To Experimentation Correlational and Quasi Experimental Design PDF

    Uploaded by

    KIM ABIGAIL ABAG
    The document discusses alternatives to experimentation for studying causation, including correlational designs, quasi-experimental designs, and causal modeling. Correlational designs examine relationships between variables but do not imply causation. Quasi-experimental designs compare groups that differ on a variable of interest but were not randomly assigned. Both approaches can suggest causal hypotheses but do not prove causation. Causal modeling techniques like regression analysis, cross-lagged panel designs, and path analysis can help establish directional relationships between variables over time.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    5 Alternatives To Experimentation Correlational and Quasi Experimental Design PDF

    Uploaded by

    KIM ABIGAIL ABAG
    44 views47 pages
    The document discusses alternatives to experimentation for studying causation, including correlational designs, quasi-experimental designs, and causal modeling. Correlational designs examine relationships between variables but do not imply causation. Quasi-experimental designs compare groups that differ on a variable of interest but were not randomly assigned. Both approaches can suggest causal hypotheses but do not prove causation. Causal modeling techniques like regression analysis, cross-lagged panel designs, and path analysis can help establish directional relationships between variables over time.

    Original Title

    5-Alternatives-to-Experimentation-Correlational-and-Quasi-Experimental-Design.pdf

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document discusses alternatives to experimentation for studying causation, including correlational designs, quasi-experimental designs, and causal modeling. Correlational designs examine relationships between variables but do not imply causation. Quasi-experimental designs compare groups that differ on a variable of interest but were not randomly assigned. Both approaches can suggest causal hypotheses but do not prove causation. Causal modeling techniques like regression analysis, cross-lagged panel designs, and path analysis can help establish directional relationships between variables over time.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    44 views47 pages

    5 Alternatives To Experimentation Correlational and Quasi Experimental Design PDF

    Uploaded by

    KIM ABIGAIL ABAG
    The document discusses alternatives to experimentation for studying causation, including correlational designs, quasi-experimental designs, and causal modeling. Correlational designs examine relationships between variables but do not imply causation. Quasi-experimental designs compare groups that differ on a variable of interest but were not randomly assigned. Both approaches can suggest causal hypotheses but do not prove causation. Causal modeling techniques like regression analysis, cross-lagged panel designs, and path analysis can help establish directional relationships between variables over time.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 47

    Alternatives to Experimentation:

    Correlational and Quasi-


    Experimental Designs

    HENRY D. PALIVINO
    Psychology Instructor
    Objectives
    • Learn more techniques that do not manipulate antecedent
    conditions: correlations, other correlational-based
    methods, and quasi-experimental designs.
    • Learn how causal models can be constructed from
    correlation-based designs
    • Understand how the results of these nonexperimental
    techniques may (and may not) be interpreted.
    Correlational Designs
    • low in manipulation of antecedent conditions
    • high in imposition of units
    Correlational Study
    • degree of relationship between two traits, behaviors or
    events
    • finding the relationship between two variables but it does
    not imply causation
    • though advanced correlational method can propose cause
    and effect relationship by developing causal models. such
    as path analysis and cross-lagged panel designs.
    • Pearson Product Moment of Correlation
    • Pearson r
    Simple Correlations
    • Relationship between pairs of scores from each subject.
    • Because of the way statistic is computed, value of
    correlational coefficient ranges from -1 and +1
    General Outcomes
    • Positive/ Direct Relationship
    • Negative/ Indirect Relationship
    • No Correlation
    Direction
    -.08 Magnitude/
    Strength
    • Correlation is represented by scatterplot/scattergram
    Positive/ Direct
    Negative/Indirect
    No Correlation
    Variables which can affect correlation
    1. Curvilinear
    2. Range Truncation
    3. Outliers
    Reminders

    Correlation does not


    imply causation!!!
    • Time Spent watching TV and Aggresive Behavior
    • We cannot say that watching violent TV causes Aggresive
    Behavior.
    1
    • Innate Aggressiveness might determine a preference for
    violent TV, not the other way around.
    2
    • Innate aggressiveness results in more exposure to TV
    violence but at the same time, the more exposure a
    person has, the more aggressive he or she becomes
    • This is know as Bidirectional causation
    3
    • Some third agent may actually be causing two behaviors
    to appear to be related.
    • This is known as third variable problem
    Summary table for 4 Possible Causal
    Directions for a Simple Correlation
    1. X Y watching increases aggresiveness
    2. Y X aggressiveness inreases watching
    3. X Y both
    4. Z
    An unmeasurable third variable
    X Y (Z) causes a child to watch more
    violent TV and causes a child to
    have higher levels of
    aggresiveness
    Coefficient of Determination
    • denotes by r2
    • Estimates the amount of variablity in scores on one
    variable that can be explained by the other variable.
    • an estimate of the strength of relationship between them.
    • r=.51 r2= .26
    • 26% of the change in Y is accounted in the change in X
    • According to Cohen, r2 which is greater than or equal to
    0.25 can be considered as strong association.
    Linear Regression Analysis
    • When two variables are strongly related, the researcher
    can estimate a score on one of the measured behaviors
    from a score on the other.
    Multiple Correlation
    • a measured behavior can be predicted by a number of
    other measured behaviors rather than a single predictor
    • Multiple Correlation does not explain why three measures
    are related but it suggests the hypothesis that other
    variable is an important 3rd variable that could be
    explored in subsequent research
    • It can be supported by partial correlation
    Multiple Regression
    • 2 or more related behaviors re being correlated
    • use to predict the score on one behavior from scores on
    the others.
    Activity
    • Practicum in using SPSS
    • simple linear correlation
    • Linear Regression
    • Multiple Regression
    Causal Modeling
    Path Analysis
    -a correlation based research method that can be used
    when subjects are measured on several related behaviors.
    -model of possible causal sequences.
    -it generates important information for prediction and can
    generate experimental hypothesis
    .64 .65 .70
    Empathy Self-reported Attitude Toward Helping the
    Manipulation Empathy the Group Group
    Cross-Lagged Panel Design
    • it uses relationship measured over time to suggest the
    causal path.
    • subjects are measured at two separate points in time on
    the same pair of related behavior or characteristics
    Time Spent r=.14 Time Spent
    watching TV watching TV
    at age 3 at age 8

    r=.20 r=.07 r=.05


    r=.-.59

    Size of Size of
    Vocabulary vocabulary
    at age 3 r=.41 at age 8
    Summary of Correlation and Correlation Based
    Methods
    Correlation degree of relationship between 2 variables
    Multiple degree of relationship between 3 or more variables
    Correlation
    Linear Regression use to predict scores on one variable from scores on a
    Analysis correlated variable
    Multiple Regression use to predict scores on one variable from scores on sets of
    Analysis other variables.
    Factor Analysis Determine subsets of correlated variables within a larger set of
    variables
    Cross Lagged Measures the same pair of variables at two different points in
    Panel Design time, looks at patterns of correlation across time for possible
    direction of cause and effect.
    Quasi-Experimental
    • quasi= latin word which means “seeming like”
    • compared to experimentation, it lacks manipulation of
    antecedent condition and random assignment of
    participants.
    • instead participants are selected for the different
    conditions of the study on the basis of preexisting
    characteristics.
    • use to explore the effects of different treatments on
    preexisting group of subjects or to investigate the same
    kind of naturally occuring events, characteristics and
    behaviors that we measure in correlational studies.
    Kinds of Quasi-experimental Designs
    • Ex Post Facto Studies
    • Nonequivalent Designs
    • Longitudinal Design
    • Cross sectional studies
    • Pretest/Post test Design
    Ex Post Facto Studies
    • use if your interested to study the effect of traits, behavior
    or naturally occuring events that cannot or should not be
    manipulated
    • researcher systematically examines the effects of subject
    characteristics but without actually manipulating them.
    • researcher forms treatment groups by selecting subjects
    on the basis of difference that already exist.
    Nonequivalent Design
    • one in which the researcher compares the effects of
    different treatment conditions on preexisting group of
    participants.
    • random assignment is not possible
    Longitudinal Design
    • measuring behavior of the same subjects at different point
    in time.
    Cross-sectional studies
    • just like longitudinal design but instead of tracking the
    same group over a long period of time, subjects who are
    already at different stages are compared at a single point
    in time.
    Pretest/ Post test
    • measuring peoples behavior before and after the event
    and compared these levels.
    Summary of Quasi Experimental Research
    Designs
    Quasi Experiment Investigate differences in preexisting groups of subjects, groups
    differences on some variable may be explored or different treatments
    given to preexisting groups may be compared.
    Ex post facto explore characteristics, behaviors or effects of naturally occurring events
    in preexisting groups of subjects.
    Non equivalent groups Compares the effect of different treatment conditions on preexisting
    groups of subjects.
    Longitudinal Investigates changes across time by measuring behavior of same group
    of subjects at different points in time.
    Cross Sectional Investigate changes across time by comparing groups of subjects
    already at different stages at a single point in time.
    Pretest-Posttest Explores the effects of an event (treatment) by comparing behavior
    Design before and after the even (treatment)
    ANY QUESTIONS?

    THANK YOU! ☺

    You might also like