The use of convenient sampling may result in low external validity for the study. Only university students were used, with majority of the sample size being aged 18-25. However, previous studies indicate that the elderly may be more sensitive to the beneficial effects of caffeine, suggesting that the extent of caffeine’s influence may vary with age (Swift & Tiplady, 1988; Jarvis, 1993). As such, using only student volunteers means that the results of this experiment cannot be fully generalised to the larger population.The differences between groups may be partly due to caffeine withdrawal effects. To eliminate systemic caffeine, subjects are requested to abstain from consuming caffeinated drinks for 6-8 hours. This may have lead to caffeine withdrawal, where cognitive abilities are adversely affected (Juliano & Griffiths 2004). Unlike the placebo group which did not consume caffeine, the experimental group’s caffeine intake may have relieved withdrawal symptoms. Therefore, the relatively superior performance of the experimental group may be due to the effects of caffeine withdrawal rather than caffeine itself.The use of white noise may be a stressor rather than a distractor, causing anxiety in the distraction group (Beutel et al., 2016; Haines, Stansfeld, Job, Berglund, & Head, 2001). At high levels of anxiety, attention is adversely affected (Waters, Nitz, Craske, & Johnson, 2007). This may have a confounding effect on results, where attention is affected not due to noise as a distraction, but as a source of anxiety (Derryberry, & Reed 2002). It is recommended that future studies use a more generalizable sample selection method such as random sampling. The confounding effects of caffeine withdrawal cannot be discounted from caffeine experiments, with 90% of adults in the United States consuming caffeine regularly (Frary, Johnson, & Wang, 2005). This means that a large proportion of the population is vulnerable to caffeine withdrawal. To gain more insight on the relationship between systemic caffeine elimination and withdrawal symptoms, perhaps future studies could track systemic caffeine levels with saliva samples (Christopher, Sutherland, & Smith, 2005) while testing cognitive abilities. If so, it may be that the confounding effects of caffeine withdrawal can be accounted for in future experiments. As for the anxiety-inducing effects of white noise, perhaps both pre-test and post-test questionnaires can be done to measure the overall effects of anxiety from sound on attention.