Poor computational fluency is a signature characteristic of mathematics learning disabilities in elementary school (e.g., Geary, 2004; Hasselbring, Goin, and Bransford, 1988; Jordan and Montani, 1997; Jordan, Hanich, and Kaplan, 2003a, 2003b; Ostad, 1998; Russell and Ginsburg, 1984). Computational fluency refers to accurate, efficient, and flexible computation with basic operations. Weak knowledge of facts reduces cognitive and attentional resources that are necessary for learning advanced mathematics (Goldman and Pellegrino, 1987). Computational fluency deficits can be reliably identified in the first few years of school and, if not addressed, are very persistent throughout elementary and middle school (Jordan, Hanich, and Kaplan, 2003b).

Children around the world move through a learning path of levels of solution methods for addition and subtraction problems. These levels become progressively more abstract, abbreviated, embedded, and complex. As they move through the levels, many children use a mix of strategies that vary according to number size and aspects of the problem situation (Geary and Burlinghman-Dubree, 1989; Siegler and Jenkins, 1989; Siegler and Robinson, 1982; Siegler and Shipley, 1995).

In contrast, young children with a mathematics learning disability rely on the most primitive Level 1 methods for extended periods in elementary school, do not use efficient counting procedures (e.g., counting on from the larger addend), and make frequent counting errors while learning to add and subtract (Geary, 1990). They also lag behind other children in the accuracy and linearity of their number line estimates (Geary et al., 2007). Researchers have differentiated children with a specific mathematics learning disability from those with a comorbid learning disability in both mathematics and reading. Jordan and colleagues (Hanich et al., 2001; Jordan, Hanich, and Kaplan, 2003a; Jordan, Kaplan, and Hanich, 2002) as well as other researchers (e.g., Geary, Hamson, and Hoard, 2000; Landerl, Bevan, and Butterworth, 2004) suggest that the nature of the mathematical deficits is similar for both groups, although children with the comorbid condition show lower performance overall. What differentiates children with a mathematics-only disability from those with combined mathematics and reading learning disabilities is that the former group performs better on word problems in mathematics, which depend on language comprehension as well as calculation facility. The potential for catching up in mathematics is much better for children with a mathematics-only disability, who can exploit their relative strength in general language to compensate for their deficiencies with numbers.

Some research shows that mathematics learning disabilities can be traced to early weaknesses in number, number relationships, and number