SEM Series Part 7: Building Your Structural Model

Video quality	The size	Download

Информация о SEM Series Part 7: Building Your Structural Model

Название	:	SEM Series Part 7: Building Your Structural Model
Продолжительность	:	11.21
Дата публикации	:
Просмотров	:	67 rb

Dear james, first off thanks for all the amazing videos I’ve done your SEM course on my educator and it’s been a life saver I am writing up my PhD analysis chapter now and have one massive issue: when I create my SEM model with variables imputed from my CFA, I get totally different results than when I build the model using stndardized variables computed from the item means in SPSS Is that possible? The imputed ones seem so over the top as all paths are p<0001 … just seems weird What to do?? Would really appreciate the advice
Comment from : @neanova7899

Hi James, does analysis with control variables only work with composite variable? Thanks
Comment from : @CunghoctiengAnh

Dear Sir,brthank you so much! Your videos helped me a lot to build my SEM Model and complete my Phd ResearchbrbrMoving on to my post-doc I intend to standardize some of the psycometric scales that I have created with SEMbrCan you please recommend a way for standardization and a video tutorial for this?
Comment from : @ASTROKALYMNOS

Hi James, thank you for your amazing videos You said "Dave Kenny says it is ok to covary suggestions from modification indices if they are on the same level" (I suppose that means eg not to covary error terms and latent variables, right?) My question is do you have a source from Kenny for this prodedure for citation? Thank you very much!
Comment from : @JonathanRothTV

Hello James brbrThank you very much for the amazing and helpful videos I am constructing my structural (latent ) model I have 5 IVs and 1 DV and a mediator I have 3 issues/questions to ask please:br1 Error terms for the mediator and DV are highly correlated when I run my latent structural model Fit indices are looking good but they drop when running the structural model (ie CFI drops from 0963 to 0954)br2 It looks like there is a significant path from one of the IDs to DV that emerges but is not a part of my conceptual model When I re-sun the analysis after linking that path fit improves, but all my hypotheses, which look good prior to linking the path, are completely ruinedbr3 The third question is about competing models Do I need to compare SEM models as a part of my PhD thesis? Can I continue the analysis without recognising the significant path that emerged but is not a part of theory that I am testing?brbrThank you very much and I apologize for the long comment
Comment from : @hharb8492

Dear James, do you have any video tutorial on item parceling in AMOS? Please suggest
Comment from : @sharjeelphd

Thank you very much
Comment from : @prashantkulkarni8314

Hello Jems ,I would like to know if there is video tutorials on how can Interpretive structural model can be validate using SEM
Comment from : @prashantkulkarni8314

Hi James, Thank you so much for the nice video According to the model above, when we go through the results of the association between joy and infoAcq, whether joy was adjusted for the computer use, play and atypuse?
Comment from : @shumintang119

Dear James, Hope this message finds you wellbrbrcloud you please let me know how to handle the following message in AMOS:brbr"It is estimated that the predictors of Self_Efficacy explain 0 percent of its variance In other words, the error variance of Self_Efficacy is approximately 100 percent of the variance of Self_Efficacy itself"brbrafter I run the SEM all values look fine except that i have 000 "squared multiple correlations" over 2 reflective variablesbrbrthank you
Comment from : @hamzahboul9544

Hi, Jamesbrhow to build the structural model from the measurement modelbrHow do we take the values of Latent variablesbrCould you please tell me if you have any video for this
Comment from : @santoshbasavaraj6738

hi Mr Gaskinbri am new to SEM i am really confused why are constructs drawn using ellipse while making the measurement model become recntangles and squares while making the structural model please help me
Comment from : @jannethalcantara5742

Thanks for the awesome vids! In a paper I am writing I used the full latent model for CFA and composite model for SEM path analysis Is this okay? I did model fit for both rounds I am being told I cannot (apples and oranges) and if I do I need evidence that I can ??? I couldn't find papers to discredit this method or otherwise
Comment from : @ColinL08

Hello James - just one quick question In my structural model there is a conflict between CFI (which is very good, 0949) and RMSEA (which is poor, 0353) Also, the SRMR is pretty good (006) Is the model acceptable? What to do about that RMSEA? Can this conflict be explained? Thanks a million!!
Comment from : @BogdanAnastasiei

Amazing work James
Comment from : @santoshbasavaraj6738

Thank you for your videos
Comment from : @spss8985

Hi James, I would like to know if there is a video tutorial on how to create those composite variables that you used in this video? Cause right now I have a bunch of observed variables that represents a few different latent variables I would like to create those latent variables and run my analysis with it but I'm not sure how I can do that brThank you
Comment from : @lawjiawern5172

You do not have any thing about categorical variables in SEM, my question is about model fit in such model, could we depend on confidence interval to judge significance?
Comment from : @alialmousawi6451

Hi JamesbrbrWhy is it we covariance the independent variables? It seems like in my model that the variables are ”stealing” from each other, which doesn’t provide such great result, as of the covariance weren’t drawn?
Comment from : @Lagkage89

Hi James can the error terms on the DVs of the latent variables be covarried if Amos suggests so
Comment from : @joanitakataike3828

Hi James I have a problembrWhen I tried run the model, AMOS give me a error, and say this:brbr"An error ocurred while attempting to fit the model In order to analyse data with missing observations, you must explicity estimate means and intercepts In amos Graphics, place check mark next to "Estimate means and intercepts" in the "Analysis Properties" window In programming the Amos engine, use the ModelMeansAndIntercepts method"brbrI already check my data base for empty values, but all they are greatbrbrI can't find the error in my model brbrCan you help me please?brbrThanks :D
Comment from : @CarlaliS9

Dear Dr James, I am analyzing the moderation I built the structural model in two ways like your video and the result of the first way is much better in model fit than second way (use composite variables) CMIN/Df of second way is around 15, so high Could you please explain why it is different?brAnd another question is in the structural model, the arrow of moderator (Alternative attraction) will go from that moderator to dependent variable (Intention) and to independent (Attitude) or only from moderator to dependent br AT <-- AA --> INTbrbrThank you so much for your support
Comment from : @VanNguyen-jo9gm

Hello Mr Gaskin I have a question that I can connect e terms of different factor with same level as you said in this clip?
Comment from : @minhmai3954

Hi James Thx again for your videobrFor my Thesis I am gonna to analyse the effect of organizational culture on organizational performance I finished my measurement model and did some related tests like Cronbach's alpha, AVE, Loading etcbrNow I need to build the structural model and have some problems Don't exactly know, how to build itbrbrIf you don't mind, I would like to send you my current measurement model so you can see itbrbrBest,brFabio
Comment from : @fabiosestito6917

Hi DrGaskin, I have a question that when I have CLF in my model and I run model fit with CLF, I found that SRMR didn't show its result What should I do if I want an SRMR in for the model fit? Thank you so much
Comment from : @jwachira

Hi, James Thank you very much for your wonderful lecture series! They really help me a lot I met a question when performing SEM analysis in AMOS My DV and Most of the IVs are latent variable But one IV is a single observed variable When I draw an arrow pointing from that IV to DV, the performance seems to be problematic Also, I can't draw co-variance between this IV and other independent latent variables Could you give me some suggestions? Whether it is allowed to have both observed and latent variables at the same level? If not, what can I do to fix it? Thank you very much for your attention!
Comment from : @lethecat

Hello James, i think i dont have to mention again how great your viedos are (but they are really helpful :) ) brbrIn this video you talk about the modification indices and why they are just allowed between error terms and on the same level You mentioned you do this because Dave Kenny says this I have read now 2 papers of Kenny but i dont find him talking about that Can you please tell me the paper ?
Comment from : @SuperQ76

Hi James, Thanks for your excellent videos I learnt lot from them Still I have a doubt How can I confirm our model is fit? what are the specific measurement should be checked?
Comment from : @ishansenarathna4626

Hi James, Thanks a lot for providing these lecture series which are very helpful I have a question about applying sampling weight when performing SEM analysis in AMOS I know SPSS allows users to apply sampling weight but I wonder how can I do it in AMOS? Thank you very much
Comment from : @LeonaFavoritetLove

Hi James, Thanks a million for these invaluable seriesbrMy question is how did you reduce number of observed variables (ex joy1, joy2,etc) into smaller number of latent variables(ex joy) Is there any recommended method?brThanks a lot
Comment from : @mosalemksa

Hi, James, I need a little help I have fixed four of the paths to be 1 in order to run the SEM analyses and now I don't know how to see whether these four paths are significant, can you please help me
Comment from : @nikolinarizanovska9731

Dear Dr Gaskin, I have second order constructs in my model and I can not run the SEM with them I checked your videos and I think they are related to first order constructs Please let me know whether I can run a model with second order constructs in SEM And if it is doable, I do appreciate if you upload a SEM model with second order or third order constructs Thanks
Comment from : @arashzadeh1600

Dear prof James, Im aza from Malaysia how to interpreting the data for path analysis if the model show 1000
Comment from : @marnizahrahim7976

Dear James, I have run the structural model using both hybrid and composite but the result is different why? when I use composite, Amos cannot create goodness of fit validation, while in hybrid structural model it gave GOF But I saw in this video you have the gof (cmin/df, etc) Thanks in advance
Comment from : @firmanwidya8899

Hi James, I am trying to run a model in which I have index variables represented by rectangulars and latent variables represented by elipses However, AMOS doesn't let me draw the path arrows from the index variables to the latent variables Could you please help me with this problem?brbrThank you! brbrBoryana
Comment from : @boryanadimitrova3154

Dear James, Thank you so much for your videos, I have 3 variables (RB which has mediated between SP and SAT) and they used 5-point Likert scale but I should add the other variable in this study as in-dependent variable that is Fitness level which has three categories High, mediate and low level fitness Is it possible to put fitness level as in-dependent because the scales are different? If Not, could you please tell me how can participate this variable in this framework?
Comment from : @mahdiesfahani4275

Hi James, Your videos have been very helpful and are easy to understand- thanks for that But I keep getting an error message as I try to run my structural model It's telling me that the sample moment matrix is not positive definite and gives a list of reasons why this may be I don't think any of the reasons would apply in my case but when I click 'allow non-positive definite sample covariances' as they suggest as a work around there are no model fit indices (AGFI, CFI etc) Thanks
Comment from : @drjessicavredenburg

Thank you!
Comment from : @jenniferzarzosa

Dear James, Firstly, thank you for sharing all this useful SEM info I have created a structural model, but it does not have good model fit The modification indices suggest to covary an error term with a factor, which I don't think is possible Another suggest is to create a path from IV to DV Is that possible? If the model still has poor model fit, does this mean that I cannot go to the next step (SEM Series Part 8: Mediation)?
Comment from : @jenniferzarzosa

Wow thanks! I'm doing this analysis for my Master Thesis You already have a mental share in it if the analysis gets any good ;)
Comment from : @philippreineke7172

I have never tested a non-recursive model in AMOS, although I know you can do it I'm not sure what the caveats of such an analysis are Here is a paper that discusses it though: "Testing Reciprocal Relations by Nonrecursive Structural equation Models Using Cross-Sectional Data" Best of luck
Comment from : @Gaskination

Dear James, Thanks a lot for your videos! They have been incredibly helpful in discovering SEM I just discovered that my model is a "non-recursive model" with a feedback loop (A --> B --> C --> A) I was wondering whether such a model can be modelled in AMOS as shown in your video (using cross-sectional data) or if there are special issues I need to pay attention to
Comment from : @philippreineke7172

I created composites during the CFA There is a video about this called "Imputing composites variables in Amos"
Comment from : @Gaskination

Dear James! Your videos have been great help for me as I am doing my own SEM However in this video it is unclear for me, how you have created some variables When you do structural model second time, I dont know, how your variables like Play or Usefulness have been made I have watched all your videos in SEM series, but I dont find answer there I guess sum variable is not good enaugh for this kind of analysis
Comment from : @markohavukainen909

Dear James, Is there any way to use a binary latent variable (ie Adoption) in the final structural model? My final DP is the adoption and I prefer to use a binary construct rather than measuring the extent of adoption But I am not sure if I can include it in Amos analysis I really appreciate your help
Comment from : @kiakashi2125

It was perfect Thanks so much
Comment from : @mahdiesfahani4275

Check out my video about handling 2nd order factors in amos
Comment from : @Gaskination

Hi dear James I really enjoy when I follow these videos I have a big framework with 8 constructs but 5 of my constructs have dimensions What should I do with these dimensions to make a model in AMOS? For example one of the constructs has five dimensions and every dimension has 5 questions the other has 4 dimensions with 5 items each and etc Thank so much for your help
Comment from : @mahdiesfahani4275

Thank you so much !
Comment from : @gustavoalmeida8344

continued - fit index and AIC, EICV , indicate the simpler model is best Is that so ?? If the retained components are as good to explain the data should I say that the simpler theory predicts the data better than the other ? And if so, can I drop the other component (latent variable) and say that online attitude predicts INTENTION (also theoretically sound) ? Thanks a million !
Comment from : @gustavoalmeida8344

Professor Gaskin, This video series and all your material are a life-saving device ! I have a doubt though I have built a SEM and my model fits well, and explains 50 of the variance of the dependent concept (latent) But I am using theory of planned behavior, with 3 components(atitude,personal control,subjetive norms(sn)) There's a simpler model with 2 components(attitude,sn) When I drop the personal control component the fit improves and model still predicts well (and comparative -->
Comment from : @gustavoalmeida8344

The CFA is for testing the factor structure, the validity and reliability of the latent factors The full SEM model uses the same structure, but removes covariances among the endogenous variables and adds regression lines This model takes into full account the measurement error of the latent factors The path model is a simplified version of the SEM and ignores some error, but also greatly simplifies the analysis The results will not differ greatly
Comment from : @Gaskination

Would you mind to explain a bit in detail what is the real difference between the two methods, ie, SEM model and path analysis model? also confirmatory factor analysis model (Shah & Goldstein, 2006, pp150) In my case, using path analysis model, all paths coefficients but in using SEM model some are not Why we use two methods and how to justified it? please
Comment from : @nayabeyes