27.12.2013 Views

The influence of the place-value structure of the Arabic number ...

The influence of the place-value structure of the Arabic number ...

The influence of the place-value structure of the Arabic number ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

problems, respectively. This was surprising and needs fur<strong>the</strong>r explanation because usually<br />

WM capacity in negatively correlated with addition performance in adults (e.g., Imbo et al,<br />

2007c).<br />

In our view <strong>the</strong>re are at least two possible explanations accounting for <strong>the</strong>se<br />

unexpected results. (i) Zuber et al. (2009) observed transcoding performance to be reliably<br />

moderated by WM capacity in a sample <strong>of</strong> first graders from which <strong>the</strong> sample <strong>of</strong> <strong>the</strong> present<br />

study was recruited. <strong>The</strong>refore, possible <strong>influence</strong>s <strong>of</strong> WM capacity on addition performance<br />

in <strong>the</strong> present study may have been driven by its interrelation with transcoding performance.<br />

This interpretation is backed by <strong>the</strong> current results as a measure <strong>of</strong> first grade WM span was<br />

incorporated in <strong>the</strong> regression model predicting <strong>the</strong> carry effect (i.e., CE) and <strong>the</strong> model<br />

accounting for performance for carry addition problems (i.e., visuo-spatial WM span). In this<br />

context, it may be important to note that <strong>the</strong> raw visuo-spatial correlation with later addition<br />

performance is not significant and that visuo-spatial WM only became a predictor due to its<br />

negative correlation with pure inversion errors in transcoding: Thus, better performance in<br />

first grade transcoding (i.e., fewer errors) is associated with better WM (see also Camos,<br />

2008; Zuber et al., 2009). So, <strong>the</strong> beneficial <strong>influence</strong>s <strong>of</strong> WM may already be partially<br />

considered by <strong>the</strong> variance explained by <strong>the</strong> numerical predictors. <strong>The</strong> remaining <strong>influence</strong><br />

may be attributed to WM serving as a suppressor variable; by suppressing irrelevant variance<br />

<strong>of</strong> <strong>the</strong> numerical predictors this resulted in a higher overall R 2 . (ii) Ano<strong>the</strong>r possible reason<br />

why we did not obtain reliable <strong>influence</strong>s <strong>of</strong> WM on addition performance may be that unlike<br />

Imbo et al. (2007a) <strong>the</strong> addition task employed in <strong>the</strong> current study required choosing <strong>the</strong><br />

correct result from two presented solution probes ra<strong>the</strong>r than actively producing <strong>the</strong> correct<br />

result. Hence, partial results necessary to be kept in mind when trying to compute <strong>the</strong> result<br />

could have been verified by comparing unit and/or decade digit <strong>of</strong> a preliminary result to <strong>the</strong><br />

corresponding digits <strong>of</strong> <strong>the</strong> solution probes. However, it may have been <strong>the</strong> case that children<br />

with a good WM might not have tried to bypass <strong>the</strong> addition task by checking <strong>the</strong> plausibility<br />

164

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!