3003PSY Week 11 Mini-Transcript PDF

SPEAKER 0 Welcome back to the mini later. Siri's 33 PS Y Dr Natalie Knox In In. SPEAKER 1 In this mini lecture, we continue to look at longitudinal desires this time residual ist magnitude nor regressions to just limitations of simplex design and longitudinal correlations. SPEAKER 0 Researchers. K...

SPEAKER 0 Welcome back to the mini later. Siri's 33 PS Y Dr Natalie Knox In In. SPEAKER 1 In this mini lecture, we continue to look at longitudinal desires this time residual ist magnitude nor regressions to just limitations of simplex design and longitudinal correlations. SPEAKER 0 Researchers. Khun Ridiculous. The Dependent Variable. This is by running a multiple regression with the dependent variable measure to time to entered as outcome variable, and both the DV measured it. Time one and the ivy tented as the predictor. SPEAKER 1 By now, we know that using regression that we can look at, the parents explained in a criterion veritable by SPEAKER 0 the predictive variables. More importantly, that we can petition variance explained as a result of entering the dependent variable a time one with together with the we can account for. The correlation between the variable across the two time point, in other words, establishes stability off the dependent variable, and then we can remove the shared variance in the dependent variable across the time points from the rest of your analysis. This process allows researchers to predict change. SPEAKER 1 This is because by entering the dependent variable a tight SPEAKER 0 one. We only have unique variant of remaining in the dependent Variable time tio. The unique variant left is the residual various not explained by the sand construct measure earlier, and this reflects the changing a constructor of the time. Let's look at this visually again. SPEAKER 1 This is a simplex design, the last test for stability SPEAKER 0 and change. Recall that the non shed variance between the dependent variable SPEAKER 1 time to that is not due to stability. SPEAKER 0 It with the construct by one, represents change. Here is longer tube no correlation model that looks at the association between the Ivy, a time one and the deviate Titou, the residual lies, longitudinal regression. SPEAKER 1 It's essentially a combination of the simplest design and the SPEAKER 0 longitudinal correlation. SPEAKER 1 You see that by measuring the dependent variable twice that SPEAKER 0 when we look at the red circle, the dependent variable in time to the various explained by their dependent variable time. What has been partial doubt? This is the part due to stability, the overlap with the blue circle and also you can see that the overlap with the idea, the deviate time one has also been removed. This is that she had variance with the green circle. The remaining inexperience between the ivy and the dependent variable. A time, too, is the region around shaded area. This's the residual of the dependent variable and therefore, is how much the ivy, a type one predicts the change in the dependant bearable. Okay, that's a little bit brain melty. Let's look at it. And the example. This is studied by doctor Alex MacDonald, who was a previous convener of this course in this study here. Disco Walls has recruited 374 high school students who are followed over two years in grade nine and 10. They were interested in whether connecting with the community drink extracurricular activities was related to hire wellbeing. This is a conceptual residual Is regression model being well trained researchers? They knew that well being is a well known stable construct, so there should be fairly high stability. They also knew that there should be a high cross sectional relation between feeling connected to the committee and higher wellbeing. Theoretically, they proposed that feeling connected to the community should cause higher wellbeing. Over time, the stability of wellbeing indicated it shares the same bearing. So the time the cross sectional relationship between the over and Devi indicates that feeling connected to the community shares variance of well being. But how do we know that feeling connected with the committee predicts change and well being and not stability? To test this they ran a residual is regression. What they found was that by controlling for the previous year's wellbeing, they're feeling connected with the committee makes the change in the adolescents wellbeing and that by controlling for well being a time won the stability and the construct had been removed, allowing a prediction of change the beer weight from links to the community. It time one was a predicted. It was a significant victor of the residual of well being at a title. As a result, they concluded that links to the computer predicts the change and well being again, looking at the Bend diagram that links to the committee account of uniquely for well being a time, too, after controlling for well, being a Taiwan. The inexperience after controlling for the dependent variable in Taiwan and the overlap between the either the DV is that time one is a residual, as always, their strength and weaknesses to this approach on the streets of this approach that the correlations between the variables that Taiwan are extinct throat, for We also know that the stability within the Devi is accounted for, and consequently, this analytic technique allows a researcher to predict change. In other words, find a longitudinal effect. However, one of the problems here is that Temple President is still not identifiable because there's no test for by direction relationships. In the next many letter, we look at the cross legged model, which allows us to test for both beauty and by their models. SPEAKER 1 In summary, residual is longer. General regression combines the strength of the simplex design and the longitudinal correlation, and this designed the dependent variable measure that time one, isn't it into the analysis for along with the I've to predict the defendant variable measure the time, too. This partitions of variance and independent variable measure that time too, in the very shared with the dependent variable measure time, one which represents stability and the unique variant or the residual experience, which is now at the change. The independent variable can then be used to predict the change independent variable that time, too. This approach is limited in testing on foot temporal presidents, and they cannot test bi directional relationships