In order to analyze, understand, and remember spoken language, multiple complex cognitive processes are involved. The field of cognitive psychology aims to quantify performance in such settings regarding, e.g., comprehension, listening effort, or verbal short-term memory. A well-established paradigm to examine short-term memory performance is the serial recall task. In this paradigm a list of items, e.g., numbers or short words, is presented to the participants. This list has to be repeated by the participants in the correct order. Past research has focused to a considerable extend on the effect of altering the experiment design on the participants' performance, e.g., with regard to input or output modalities. However, studying the influence of different acoustic conditions has mostly been limited to adding different types of background noise to the auditory presentation of the item list. For this talk, a pilot study was carried out examining the influence of varying reproduction methods on performance in the serial recall task. This includes spatial static solutions, auralization with plausible binaural cues by introducing individual head-related transfer functions (HRTFs), as well as commercial out-of-the-box reproduction systems. The results can not only help understanding cognitive speech processing but also improve close-to-reality design of acoustic virtual environments.