Hey there! Sign in to join this conversationNew here? Join for free
    • Thread Starter
    Offline

    0
    ReputationRep:
    Hi all,

    I'm reading an article at the moment and there seems to be an independent variable (experimentally manipulated) and a DV. But there also seems to be a moderating variable which is not directly manipulated by the experimenter, which is a personality variable called cognitive avoidance.

    I was just wondering what the moderating variable would be called? Is it a 'quasi-experimental variable?'

    Thanks! x
    Offline

    0
    ReputationRep:
    Moderator?

    Covariate?
    • Thread Starter
    Offline

    0
    ReputationRep:
    Yeah, i thought it might be a moderator. But the authors seem to have used a multiple regression to analyse the interaction between cognitive avoidance and the IV.....Can you use multiple regression on quasi-experimental variables?
    Offline

    0
    ReputationRep:
    If the IV is essentially something that the participants walked into the experiment with, e.g. gender or a pre-existing level of cognitive avoidance, then it is quasi-experimental, yes.

    You can use DVs and regression to predict some quasi-experimental variables.
    Offline

    0
    ReputationRep:
    (Original post by BouncyBabe5)
    Yeah, i thought it might be a moderator. But the authors seem to have used a multiple regression to analyse the interaction between cognitive avoidance and the IV.....Can you use multiple regression on quasi-experimental variables?
    regression can readily handle interactions of many forms (cont*cat, cat*cat, cont*cont, 3-way etc). Indeed, the ANOVA is merely a form of regression.

    I would think that cognitive avoidance is a continuous predictor, while your 'IV' is a categorical predictor. Thus the interaction is continuous*categorical.

    I assume the idea is that your IV (categorical predictor) moderates the relationship between cognitive avoidance and your outcome variable (i.e., dependent). ABE: in sum, the Cognitive Avoidance vs. Outcome slopes vary across the levels of the categorical predictor (i.e., your IV).

    I don't think terms like 'quasi-experimental' are very helpful (nor informative).
    • Thread Starter
    Offline

    0
    ReputationRep:
    (Original post by Psych!)
    regression can readily handle interactions of many forms (cont*cat, cat*cat, cont*cont, 3-way etc). Indeed, the ANOVA is merely a form of regression.

    I would think that cognitive avoidance is a continuous predictor, while your 'IV' is a categorical predictor. Thus the interaction is continuous*categorical.

    I assume the idea is that your IV (categorical predictor) moderates the relationship between cognitive avoidance and your outcome variable (i.e., dependent). ABE: in sum, the Cognitive Avoidance vs. Outcome slopes vary across the levels of the categorical predictor (i.e., your IV).

    I don't think terms like 'quasi-experimental' are very helpful (nor informative).
    Thanks for your help!
    It seems that the authors have measured attitudes to X before the manipulation, manipulated the IV, and then measured attitudes afterwards. They then seem to use the change in scores (between pre- and post-test) in their regression, as the DV. Is this appropriate? I wasn't aware you could use a change in scores for a DV in multiple regression?

    Sorry if this is confusing!
    Offline

    0
    ReputationRep:
    (Original post by BouncyBabe5)
    Thanks for your help!
    It seems that the authors have measured attitudes to X before the manipulation, manipulated the IV, and then measured attitudes afterwards. They then seem to use the change in scores (between pre- and post-test) in their regression, as the DV. Is this appropriate? I wasn't aware you could use a change in scores for a DV in multiple regression?

    Sorry if this is confusing!
    Yeah, no reason why one couldn't. As long as the outcome variable is a continuous variable then the regression would be fine (along with a few other assumptions).

    The design actually sounds much like a pre-test/post-test non-equivalent ANCOVA (where the pre-test is used as a covariate). Thus, by using the pre-post attitude approach, they control for individual differences (and any issues with allocation to groups).

    If you are looking for problems here, then spurious effects from regression to the mean could be one for the pre/post setting (would apply to both the ANCOVA and regression pre/post case). This is when random assignment (which is where the true vs. quasi experimental design issue makes more sense) would be helpful.
    • Thread Starter
    Offline

    0
    ReputationRep:
    (Original post by Psych!)
    Yeah, no reason why one couldn't. As long as the outcome variable is a continuous variable then the regression would be fine (along with a few other assumptions).

    The design actually sounds much like a pre-test/post-test non-equivalent ANCOVA (where the pre-test is used as a covariate). Thus, by using the pre-post attitude approach, they control for individual differences (and any issues with allocation to groups).

    If you are looking for problems here, then spurious effects from regression to the mean could be one for the pre/post setting (would apply to both the ANCOVA and regression pre/post case). This is when random assignment (which is where the true vs. quasi experimental design issue makes more sense) would be helpful.
    OK, thanks again! It makes more sense now! x
 
 
 
  • See more of what you like on The Student Room

    You can personalise what you see on TSR. Tell us a little about yourself to get started.

  • Poll
    Will you be richer or poorer than your parents?
  • See more of what you like on The Student Room

    You can personalise what you see on TSR. Tell us a little about yourself to get started.

  • The Student Room, Get Revising and Marked by Teachers are trading names of The Student Room Group Ltd.

    Register Number: 04666380 (England and Wales), VAT No. 806 8067 22 Registered Office: International House, Queens Road, Brighton, BN1 3XE

    Quick reply
    Reputation gems: You get these gems as you gain rep from other members for making good contributions and giving helpful advice.