in this video i'm going to show you how to start a new project with smart pls4 so as you might suspect go to the top left and click on new project you could also go to files and select new project now there are three other types of specific projects which we could also select i'll look at those in more detail in a different video for now i'm just going to click on the icon new project he wants me to name the project youtube tutorials hit create and now you'll see on the left we have a new project called youtube tutorials let me zoom in and the first thing we need to do is connect a data set to this project and then create a model so if i click on import data file i can go explore for that data i have some data in my downloads folder what's pretty cool about smart pls4 is that it can handle many different file types csv txt sav from spss xls from excel way better than in the past when we can only use dot data.csv files so i'm going to go down and select an spss file open it's importing it it's doing a careful screening of the data one cool new feature about smart pls4 is it now gives a type of scale to every variable so for example on id i could say that this is metric ordinal or categorical this is just an id so i'll keep it metric i'm not actually going to do any analyses with id but i happen to know that q1 and q2 etc are ordinal so i can select ordinal let me speed this up well that took forever i'm guessing in the future there will be a shortcut or some better guessing on the part of the software so we're just left with age which is metric it's just actual years in age gender which is listed as binary marital status which is categorical and people in the household which is metric years in the job metric customer interactions metric years in the firm metric education income education is categorical income is met is actually ordinal and job category as binary okay now those are all done you can see this big yellow warning down below it says we've found missing values in fact we found 22 missing values we're going to assume those empty values are missing if in addition you have a missing marker value sometimes people put in something like negative 999 as a missing marker then you can add that here i don't have a missing marker but if i did i would put it right here and then hit apply marker all right then go to import and here is the view of your data notice it does some basic descriptive statistics on your data tells you the type that you specified how many are missing for that variable the mean median the minimum the maximum that were possible on that scale but then the actual minimum and actual maximum that were observed standard deviation kurtosis skewness and the kramervan missy's p-value which is just a test of distributional normality which i've never actually used but i'm guessing is a lot like other normality tests where a p-value less than your desired threshold such as .05 or 0.1 means that it is departing from normality however it looks like all of my p-values are less than .05 and so i'm guessing this is pretty strict just like the k s or s and w tests regardless here's a view of your data and if you want to see how all of these variables are correlated you can actually just click on indicator correlations and it produces a massive correlation matrix which if you wanted you could copy out to excel later on i'll show you how to add groups or generate groups for a multi-group analysis for now i'll just show you one more thing here and that is if you think you made a mistake in your setup of the data you can click on setup and you can go back and change any of these values and the min and max values or apply your marker again i'm just going to hit cancel because i didn't make any changes and then i'm going to hit back to go back to my workspace just check that this data set is green if it is green that means it is ready to go if it is not green click on it to find out what might the problem be in most cases this should be green okay next i'm going to create a model and now i need to select a model type let me zoom in here the model types are pls sem regression or process like the process macro for mediation in spss for now i'm going to select pls sem and i will name this youtube model and save zoom out and now you get your first look at the new interface for model design for smart pls4 as before we have our selector arrow latent variable tool connector tool and then some new things like a quadratic effect comments and the gaussian copula i'm not actually sure what that is yet so let's go create a basic model i'm going to select a few variables over here on the left such as satisfaction with work i'm just going to hold shift and click the next one drag those out here and it wants me to name these so i'll name it sat w for satisfaction with work hit enter and it automatically creates a latent variable with those items let's bring out burnout for management name that burn m here's that notice it says there's some problems i know you can't see that's too small but right now the problem is this latent variable is not connected to any other latent variable and so it won't run as part of a model let's just add one more variable real quick how about ethical concerns ec1 to ec5 bring those out call that ethical concerns and now let's connect these just as before we would click on the connect option and then drag from one variable to another let's see i think ethical concerns will lead to burnout and i think that burnout will lead to a drop in satisfaction with work there's probably also a direct link between ethical concerns and satisfaction with work now if i want to move these around i need to go back to the selector tool as before in smart pls 3 and drag these around like so the last thing we'll do in this video is go to save and calculate let me zoom in and you can see there are several options we can use our standard pls sem algorithm the consistent algorithms are also still available down here for models with all reflective constructs for now i'll just use the regular pls algorithm let me zoom out and zoom back in basic settings as we had in smart pls three with path or factor right now just use path what type of results do i want to see we now have the option to see standardized unstandardized or mean centered previously we only received standardized i'll keep it as standardized for now initial waiting scheme just the default is fine and start calculation there we go it outputs standardized regression weights on these paths just as before it gives us the r squared for endogenous factors and if you hover over a variable it lists important information such as the r square r square adjusted chromebox alpha composite reliability row a and average variance extracted very handy well that's it for this video in subsequent videos i'll show you some of the other options for analysis