Linguistic and Cultural Knowledge as Prequisites to Learning Professional Translation

2.1.4. The Ex Post Facto Design

The ex post facto method may be implemented through two different designs. The one that fits this study is referred to as the criterion group study. The design of this type of research (see figure 1) consists in choosing two groups of subjects (G1 and G2). The dependent variable(s) should be present in one group and absent from the other. Of course, in the case of this investigation, we should talk about the degree of presence rather than absolute presence or absence of the dependent variable. Then the groups are compared in terms of the hypothesised independent variable(s) (X), which had already occurred.

G1 X G2 Past - · Present

Figurel: The ex post fact design adapted from" (Cohen & Manion, 1980).

This type of design is said to bridge "the gap between descriptive research methods on the one hand and true experimental research on the other" (Cohen & Manion, 1980, p. 144). Lack of manipulation of variables makes it belong to the range of descriptive methods, while the fact of choosing, observing and comparing two groups of subjects makes it seem like an experiment.

2.1.5. Sampang

The participants of this experiment were third year Translation students of Batna University. We have chosen third year and not second or first because we assumed that they should have attained a relatively advanced stage in the learning process. This advanced stage represents the dependent variable of the study. In addition, we have chosen third and not fourth year students because there exists no fourth year students in Batna Translation Department. The third year students represent the first class in this newly founded department.

Our sampling technique was based, as already explained, on the assumed "presence" and "absence" of the dependent variable. In other words, two groups have been chosen. Students who obtained the highest grades on the l^e semester exam of Arabic-English-Arabic translation

constituted the l^e group. Those who obtained the lowest grades constituted the r^d group.

The number of third year students is 141. The sample was formed with 44 students, which represents 31% of the population. Each group contained 22 students. As our sampling was based upon a specific criterion, control over other types of criteria, like sex and age was limited. Nevertheless, sex proportions were respected as much as possible. This is explained in the following table.

Table 1: Gender proportions in the ex post facto sample

Group10 Population Group 1 Group 2

Caegory N. % N. % N.

Male	45	31%	8	36%	5		23%	Female
96	69%	14	64%	17	77%		Total
141	100%	22	100%	22	100%

For the specific purpose of this study, most importance was given to prior knowledge regardless of its sources. Therefore, the students' academic background and age are not variables this study needs to control. Indeed these variables might contribute to their subjects' knowledge.