The verification of the central hypothesis about the results of the experiment implies a crossgroup
comparison of the scores (a thorough account of which is given in Appendix E). The
first way of looking at the completed experiment is to compare the post-training scores of
Group A with those of Group B. Thereby, knowing which group has received the better
scores will enable us to know which of the two trainings – segmental or prosodic – has had
the better effect on the pronunciation skills of the French EFL learners who participated. If
French speakers’ productions of English are better evaluated by native speakers and experts
after they received a training on prosodic features, then it means that enhancing prosody in
EFL pronunciation teaching may actually be more sensible than focusing on phonemic
aspects only. Accordingly, the training that has had a better effect on the learners’
productions will correspond to the aspect of English phonology that has more impact on
intelligibility, communication, and foreign-accentedness even, than the other, given that the
listeners had to rate the global English of the subjects.
Hypothesis 1
Our first claim about the results of the experiment is that the group that received a training
on suprasegmental features (Group B) will have obtained better scores than the segmental
group (Group A). Although prosody is mostly overlooked by EFL teachers, prosodic features
such as stress and rhythm are at least as important as phonemes and phones in the
acquisition of English pronunciation. As far as the final results are concerned, there are three
possibilities. If the hypothesis proves to be true, then the mean score of Group B will be
higher than the mean score of Group A. In other words, a prosodic training has a better effect
on production skills of French speakers than a segmental training. However, if the
hypothesis is wrong, then either Group A will have better scores than Group B, or the two
groups will be at the same level. In the former case, it is a training that is based on
segmentals that helps French learners improve their pronunciation skills more than a
training on prosody.
Hypothesis 2
The second hypothesis is that the segmental group may tend to be slightly better at words
(Hypothesis 2 #1), since the segmental training and its oral practice focused on the wordlevel.
However, one should not exclude the possibility that a prosodic training may help L2
speakers improve at the segmental level as well, as is maintained by Birdsong (2003). This
would imply that the prosodic group would be better at word reading. In addition, we claim
that the prosodic group will have obtained better scores in phrases (Hypothesis 2 #2), as the
prosodic training and oral practice was mainly at the phrase-level.