Rethinking Convenience Sampling: Defining Quality Criteria

Abstract

Convenience sampling is one of the most commonly used sampling procedures in second language acquisition studies, but this non-random sampling procedure suffers from a lot of problems including the inability of controlling for initial differences between experimental and control groups. The present study tries to introduce conditions and criteria which enable researchers to account for these drawbacks and at the same time make validity claims. Individual scores and group statistics are compared with regard to a group of essential factors known to be important for the purpose of the study. The overall value calculated for essential factors is then used to make judgments about the groups' comparability. The contribution of this method to the current procedures of sampling arises from its factual accuracy which is supposed to enhance the validity of findings obtained from studies employing non-probability sampling procedures.

Impact of packet loss on the perceived quality of UDP-based multimedia streaming: a study of user quality of experience in real-life environments

“…Farrokhi and Mahmoudi-Hamidabad stressed that the exclusion of outliers is especially relevant when non-probability sampling methods are used, of which convenience sampling is notorious [22]. In this study, if the subject's answers to questions A1 and A3.5 differed by more than 8 points, that questionnaire was excluded from the analysis.…”

“…This population was targeted because (a) according to Datta et al [21], video streaming services are generally used by users between the ages of 18 and 24, which corresponds with the age group of a typical student population, and (b) this population was easy accessible for conducting such a survey (i.e., the convenience sampling method [22] was used).…”

Qualitative Research in Travel Behavior Studies

“…Other method is snowball sampling, which is a technique consisting in the recruitment of subjects by other subjects already in the study because they are relatives, colleagues or acquaintances (Goodman, 1961). Convenience sample, in which the subjects are selected because of their convenient accessibility and proximity to the researcher (Farrokhi & Mahmoudi-Hamidabad, 2012).…”

Aspects de la complémentation en FLE chez des apprenants hellénophones

“…Ce genre d'échantillonnage, basé sur une sélection non-aléatoire, n'est pas représentatif d'une population plus large (inexistante dans le cas considéré ici) et les résultats obtenus ne sont forcément pas généralisables : comme le seul établissement universitaire chypriote qui offre un programme d'études françaises est celui de l'université de Chypre, l'échantillon choisi était le seul qui fût disponible. Il va de soi qu'il faut garder à l'esprit les désavantages de l'échantillonnage de convenance : possibilité de biais, présence éventuelle de valeurs aberrantes dans l'échantillon qui menacent l'homogénéité du groupe et faussent les résultats (Farrokhi & Mahmoudi-Hamidabad 2012), et que cela incite à la prudence au moment de tirer des conclusions. Cependant, bien que la randomisation et la représentativité maximale soient généralement vues comme l'idéal dans les études quantitatives, les circonstances peuvent forcer une démarche différente si la nécessité l'impose comme dans notre cas -démarche qui de plus est justifiable si, comme nous l'avons fait plus haut, on définit avec précision les circonstances dans lesquelles la recherche a été conduite (Farrokhi & Mahmoudi-Hamidabad 2012).…”