This special issue &&Best of Empirical Studies of Programmers 7'' arises, as its name suggests, out of the Seventh Workshop on Empirical Studies of Programmers, held in Alexandria, Virginia in October 1997. The series of seven workshops extends back to 1986, when Elliot Soloway and Sitharama Iyengar organized the "rst workshop to bring together researchers in what was then the still new "eld of the cognition of programming. Elliot Soloway, one of the founding "gures of the "eld, remained highly active in the organization of several succeeding meetings, later handing o! this responsibility to others who enthusiastically continue the tradition. The ESP workshops, now held every two years, are the most important gatherings of researchers on the cognition of programming held in North America. They attract an international group of researchers from North America, Europe, and Asia. The proceedings of the workshops are essential source material for researchers in the "eld.
The special issue was initiated to bring a group of high-quality papers, representing a range of issues in the cognition of programming, to a wide audience. We are very pleased to have these papers appear in the International Journal of Human-Computer Studies because of its long time role in bringing papers on the cognition of programming to the research community. Going back to the 1970s the International Journal of Man-Machine Studies, the forerunner of IJHCS, was responsible for publishing seminal work by founders of the "eld, including Thomas Green and the colleagues , 1977), Ben Shneiderman (1977), John Gould (1975), John Gannon (1976), and Ruven Brooks (1977, 1983), to name a few. The journal also played an important role in the development of the "eld when it published one of the early papers on methodology in the empirical study of programmers (Moher and Schneider, 1982). Most of the well-known contemporary researchers in the cognition of programming have published in IJHCS, which continues its important role in the "eld.
Four papers are contained in this special issue. They treat a wide range of themes in the cognition of programming and, thus, give some sense of the scope of the "eld. Petre and Blackwell's (1999) paper treats the subject of the mental imagery of expert programmers, that is, how do programmers make use of mental images in the early design of software? They "nd that the mental imagery reported has elements in common with