awareness and reading fluency factors. However, they are positively correlated with eye-

movement/reversal factors because eye-movement scores are also the time in seconds to

read numbers vertically and horizontally, and reversal scores indicate error scores.

Furthermore, all of the vision-related variables and reading fluency variables are

significantly related. Surprisingly, one of the observed variables, blending into words,

from a phonological factor has a very weak relationship with other reading-related

variables. One important fact has been detected through the correlation analysis is that

there are very high correlations between the two variables in rapid-naming and

vertical/horizontal eye movement variables. For example, DEM vertical test scores are

highly correlated with digit rapid naming and letter rapid naming (r= .849 and r= .805

respectively). Since the two constructs are significantly correlated with each other, if we

assign these scores to different constructs, it would violate one of assumptions in SEM,

multicollinearity. In some cases, multiple regression results may seem paradoxical. Even

though the overall P value is very low, all of the individual P values are high. This means

that the model fits the data well, even though none of the X variables has a statistically

significant impact on predicting Y. How is this possible? When two X variables are

highly correlated, they both convey essentially the same information. For the current

study, a rapid naming variable is very highly correlated with an eye movement variable.

In this case, neither may contribute significantly to the model after the other one is

included, but together they contribute a great deal. If you removed both variables from

the model, the fit would be much worse. So, the overall model fits the data well, but

neither X variable makes a significant contribution when it is added to the model last.

When this happens, the X variables are collinear, and the results show multicollinearity.