Guerda, OuannassaChaira, Farid2020-12-102020-12-102020http://hdl.handle.net/123456789/9548Validity is the backbone of academic research, in the sense that it ensures that the findings produced by it are valid and credible. The present study addresses two neglected types of validity in the EFL field, construct validity and statistical conclusion validity. It assesses these two types of validity in EFL master dissertations in the department of English at Larbi Ben M’hidi University. To evaluate construct validity, a number of dissertations were first selected and analyzed qualitatively in order to have a clearer picture of the situation. This was followed by a correlational study that was conducted to check the extent to which Master dissertations uphold construct validity. To carry out such evaluation, thirty first-year students were asked to answer a set of online questionnaires, namely the FLCAS test, the SSEIT test. As well as two researcher-made tests extracted from the dissertations that were chosen for analysis. The scores obtained were correlated by means of SPSS. The results indicated that the tests used by post-graduate students in their dissertations lacked construct validity, hence confirming the null hypothesis. As far as statistical conclusion validity is concerned, the procedure of NHST was applied. The results of this statistical procedure revealed that the EFL master dissertations that were analyzed uphold statistical conclusion validity to some extent, hence confirming the alternative hypothesis. Based on the findings obtained, a number of pedagogical implications were suggested to teachers and students.enConstruct validityStatistical conclusion validityOther