Fig. 3

A scree plot of a test with 24 items. The ordinate shows the weight (eigenvalue) of each factor. A test measuring one factor is characterized by a steep drop from factor 1 to factor 2. Ideally, the weight of the observed data in factor 2 falls below random simulated or resampled data. Then there is no doubt that the test measures a single factor. Most of the time, you will find that a few factors remain above the random data. This is not worrying as long as the difference is small and the elbow of the curve is clearly between factor 1 and factor 2