![]() The other components -having low quality scores- are not assumed to represent real traits underlying our 16 questions. This is because only our first 4 components have Eigenvalues of at least 1. Our 16 variables seem to measure 4 underlying factors. Select components whose Eigenvalues are at least 1.Īpplying this simple rule to the previous table answers our first research question: So what's a high Eigenvalue? A common rule of thumb is to Only components with high Eigenvalues are likely to represent real underlying factors. Each component has a quality score called an Eigenvalue. ![]() Now, with 16 input variables, PCA initially extracts 16 factors (or “components”). Factor Analysis Output I - Total Variance Explained FACTOR /VARIABLES v1 v2 v3 v4 v5 v6 v7 v8 v9 v11 v12 v13 v14 v16 v17 v20 /MISSING PAIRWISE /* IMPORTANT!*/ /PRINT INITIAL CORRELATION EXTRACTION ROTATION /FORMAT SORT BLANK(.30) /PLOT EIGEN /CRITERIA MINEIGEN(1) ITERATE(25) /EXTRACTION PC /CRITERIA ITERATE(25) /ROTATION VARIMAX /METHOD=CORRELATION. *Initial factor analysis as pasted from menu. *Show both variable names and labels in output. So let's now set our missing values and run some quick descriptive statistics with the syntax below. But in this example -fortunately- our charts all look fine. If we see something unusual in a chart, we don't easily see which variable to address. All variables have some system missing values too but the extent of missingness isn't too bad.Ī somewhat annoying flaw here is that we don't see variable names for our bar charts in the output outline.All variables have a value 8 (“ No answer”) which we need to set as a user missing value.All variables are positively coded: higher values always indicate more positive sentiments.All frequency distributions look plausible.This very minimal data check gives us quite some important insights into our data: * show variable names but not labels in output outline */ *Basic frequency tables with bar charts. set tnumbers both /* show values and value labels in output tables */ tvars both /* show variable names but not labels in output tables */ ovars names. *Show variable names, values and labels in output tables. We'll inspect the frequency distributions with corresponding bar charts for our 16 variables by running the syntax below. Now let's first make sure we have an idea of what our data basically look like. which satisfaction aspects are represented by which factors?.which questions measure similar factors?.how many factors are measured by our 16 questions?.So our research questions for this analysis are: ![]() We think these measure a smaller number of underlying satisfaction factors but we've no clue about a model. The survey included 16 questions on client satisfaction. The data thus collected are in dole-survey.sav, part of which is shown below. Research Questions and DataĪ survey was held among 388 applicants for unemployment benefits. There's different mathematical approaches to accomplishing this but the most common one is principal components analysis or PCA. The software tries to find groups of variablesĮach such group probably represents an underlying common factor. The simplest possible explanation of how it works is that That is, I'll explore the data (hence, “exploratory factor analysis”). Exploratory Factor Analysisīut what if I don't have a clue which -or even how many- factors are represented by my data? Well, in this case, I'll ask my software to suggest some model given my correlation matrix. SPSS does not include confirmatory factor analysis but those who are interested could take a look at AMOS. This is known as “ confirmatory factor analysis”. In this case, I'm trying to confirm a model by fitting it to my data. Now I could ask my software if these correlations are likely, given my theoretical factor model. Right, so after measuring questions 1 through 9 on a simple random sample of respondents, I computed this correlation matrix. So if my factor model is correct, I could expect the correlations to follow a pattern as shown below. However, questions 1 and 4 -measuring possibly unrelated traits- will not necessarily correlate. The same reasoning goes for questions 4, 5 and 6: if they really measure “the same thing” they'll probably correlate highly. Now, if questions 1, 2 and 3 all measure numeric IQ, then the Pearson correlations among these items should be substantial: respondents with high numeric IQ will typically score high on all 3 questions and reversely. For measuring these, we often try to write multiple questions that -at least partially- reflect such factors. Such “underlying factors” are often variables that are difficult to measure such as IQ, depression or extraversion. SPSS Factor Analysis – Beginners Tutorial By Ruben Geert van den Berg under Basics & Factor Analysisįactor analysis examines which underlying factors are measuredīy a (large) number of observed variables.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |