Previous evidence indicated that music can positively (Ferreri, Bigand, Bard & Bugaiska, 2015) and negatively (Iwanaga & Ito, 2002) affect encoding and retrieval of information. Moreover it was found that incongruent context at learning and test can also decrease the amount of information remembered (Godden & Baddeley, 1975). Since contrasting results exist (Coveney, Switzer, Corrigan & Redmond, 2013), the current study aimed to examine further whether memory recall would be affected when learning or testing with the music as opposed to silence. Additionally, it aimed to find out whether matching contexts at learning and test would improve the amount of information remembered. Fifty adult participants, who were friends of the researchers completed a visual task presented on the computer screen. This included memorising four lists of words, tested by immediate recall. The learning and test included background instrumental music or silence. It was hypothesised that there would be a difference between learning as well as test with the music as opposed to silence. It was also hypothesised that there would be an interaction between context at learning and test. Based on the percentage of words recalled, 2×2 repeated measures ANOVA revealed that the results did not match the predictions, therefore the hypotheses were not supported. These findings were consistent with the previous studies which showed no effect of instrumental music on memory recall (Alley & Greene, 2008) and no interaction between context at learning and test (Coveney et al., 2013). Although, there were limitations to the findings such as sample size, which could have affected the results.
Introduction
Memory is a cognitive process which includes encoding, storage and retrieval. Firstly the stimulus is collected from the environment. Then it enters short-term memory, which sometimes leads to long-term storage. Finally, retrieval takes place which is responsible for the recall of the stored information. Failure in any one of these processes may result in forgotten information. Because encoding and retrieval are integrated together, it may be difficult to distinguish which one is responsible for the lost information. Additionally, attention plays a crucial role in what is encoded, selecting only important information, therefore not all stimuli is encoded (Gillibrand, Lam & O’Donnell, 2011). Previous research has demonstrated that memory can also be affected by external factors, such as music played when learning and testing the memory. It was found that music increased encoding and retrieval of short-term memory (Ferreri et al., 2015). However, contrasting evidence exists, which showed that music interfered with memory performance (Alley & Greene, 2008; Iwanaga & Ito, 2002; Nittono, 1997; Perham & Sykora, 2012; Perham & Vizard, 2011).
The inconsistency of the findings may be explained by the nature of tasks performed as well as the type of the music presented. For example, it was shown that instrumental music disturbed comprehension of memory recall in verbal (Iwanaga & Ito, 2002) and visual tasks (Nittono, 1997) when compared to silence. However contrasting results exist, showing no difference in visual tasks between instrumental music and silence (Alley & Greene, 2008). Moreover, instrumental music interfered more with memory recall when it was fast and loud as opposed to quiet and slow (Thompson, Schellenberg & Letnic, 2011). Furthermore, vocal music or speech showed more negative effects on memory than instrumental music (Alley & Greene, 2008; Iwanaga & Ito, 2002). The study by Vredeveldt, Hitch and Baddeley (2011) revealed that impaired memory recall was caused by increased cognitive load from sensory stimuli, however the most affected was when the task and additional stimuli draw on the same cognitive processes, such as both visual or both auditory.
It was additionally found that incongruent environmental contexts at learning and test also decreased memory performance. Previous studies mostly used different rooms (Coveney et al., 2013; Isarida, Isarida & Sakari, 2012; Smith, 1982; Unsworth, Spillers & Srewer, 2012). Also, more extreme environments such as underwater and land (Godden & Baddeley, 1975) or in the air and land (Thompson, Williams, L’Esperance & Cornelius, 2001). However contrasting evidence exists, which found no difference in memory recall when tested between one room and another (Coveney et al., 2013). However the context-dependent effect in relation to music has not been investigated before.
The focus of the study was to investigate the issue of music influence on memory recall, as there are inconclusive results found in the literature. The current study, aimed to answer three research questions. Firstly, how the context of background instrumental music at learning would affect memory recall. It was predicted that there would be a difference between learning with the music as opposed to silence. Secondly, how the context of background instrumental music at test would affect memory recall. It was predicted that there would be a difference between testing with the music as opposed to silence. Thirdly, whether memory recall would be improved when the context at learning and test matched. Based on the literature review, it was hypothesised that there would be an interaction between context at learning and test.
Methods
Design
All research questions used a within-participants design. The independent variables were learning and test, each had two levels: instrumental music and silence. There were four conditions. First, learning and test with music. Second, learning in silence and test with music. Third, learning with music and test in silence. Fourth, learning and test in silence. The dependent variable for the first and second research question was the percentage of the words, which were recalled shortly after each presentation. This produced quantitative data. The words were recorded by the participant on the response sheet (see Appendix 2). Only correct responses were used. After each list was presented, participants drew a butterfly, which was implemented as a distractor and lasted 30 seconds. In order to decrease the effects of confounding variables, the order of conditions in which participants started the task at was changed, as a counterbalance. The participants completed one condition after another, without any breaks.
Participants
The participants were thirty friends of the researchers (F=14, M=16) ranging in age from 19-50 (M =30.27, SD= 10.81), who were native English speakers. This was important as the words were presented in English. Because the experiment involved visual and auditory stimuli, there was no musicians and all participants had normal or corrected to normal hearing and vision to minimise confounding variables. They participants were naïve to the hypotheses.
Materials
The lists of words were randomly divided into four PowerPoint presentations. Each included a list of ten two-syllable unrelated words (see Appendix 3), displayed for four seconds. The words were in black ink and capitals, displayed on white background in the middle of the screen. Their size was 138. Each word was used once and the order in which they were presented was randomised. The stimuli were presented on RM computer with Microsoft Windows 7. The music used was Mozart – Sonata for two pianos in D K.448 – I Allegro con spirito (https://www.youtube.com/watch?v=BuYEHFy20cE&index=1&list=LL4wqVJa04QZYCM1xk4DjH-A). The first forty seconds of the song was played during learning and testing. Headphones were used. The response sheet was an A4 sheet and a pen was provided to write the information down. These included the number of participant and counterbalance, which both were randomised, gender, age, native language, space for each condition and drawing. SPSS was used to analyse the data (see Appendix 4) using 2×2 repeated measures ANOVA to test the hypotheses.
Procedure
Participants were seated in a quiet and bright room in the researchers’ apartment in front of the computer with headphones provided. They were told that the experiment would involve memory recall of words displayed on four PowerPoint presentations including ten words on each one. It was instructed that learning and test would involve background music or silence, depending on the presentation. They were told to remember as many words as possible and after the end of each presentation they would be asked to draw a butterfly and then write the words down on the response sheet when directed.
The participants were informed that the study would take up to 30 minutes and that they would take part individually. This was implemented to minimise competition and feelings of insecurity among participants, especially as they could have known each other. It was explained that the collected data would be anonymous and confidential. The participants were given the information sheet (see Appendix 5) and consent forms (see Appendix 6), which were read and signed by all participants. They were informed that they can withdraw at any time, without providing a reason with no consequences. The learning and test took forty seconds each per presentation. After completion of the experiment the participants were debriefed (see Appendix 7) and thanked for their participation. There was no incentive given.
Results
The recorded words were transformed into percentage and transferred into SPSS. The Kolmogorov-Smirnov test revealed non-normal data, therefore three outliers were removed, because it violated assumption of ANOVA test.
The data in Table 1 shows mean and standard error of words recalled of each participant during learning and test with the music and in silence.
Table 1: The mean and standard error of the words recalled when learning and testing in silence or with the instrumental music and interaction between them.
Mean SE
Memory recall at learning (percentage)
No music 60.93 2.72
Music 58.33 2.75
Memory recall at test (percentage)
No music 60.37 2.23
Music 58.89 2.65
Interaction between the contexts (percentage)
No music * No music 63.70 3.16
No music * Music 58.15 2.97
Music * No music 57.03 2.55
Music * Music 59.63 3.45
A 2×2 repeated measures ANOVA was conducted to compare the memory recall of words in silence and with the music. As Table 1 demonstrates, learning in silence produced higher memory recall (M = 60.93, SE = 12.72) as opposed to background instrumental music (M = 58.33, SE = 2.75), p>0.05, np2 = .03, however it was non-significant. The results did not match the prediction that there would be a difference between learning with music and silence, therefore hypothesis was not supported.
Similarly, memory recall while testing in silence was higher (M = 60.37, SE = 2.23) than with the music (M = 58.89, SE = 2.65), p>.05, np2 = .03 but non-significantly. The results did not match the prediction that there would be a difference between test with music and silence, therefore hypothesis was not supported.
Lastly, there was an interaction between context at learning and context at test. Congruent contexts at learning and test produced higher memory recall (M=63.70, SE=3.16; M=59.63, SE=3.45), as opposed to incongruent (M=58.15, SE=2.97; M=57.03, SE=2.55), however it was non-significant p>.05, np2 = .12. Therefore hypothesis was not supported, since the results did not match the prediction that the context at learning and context at test would interact.
Discussion
The study examined the influence of instrumental music at learning and test on the percentage of memory recall and whether matching context at learning and test would affect it. The results revealed no significant difference between two contexts at either learning or test. Therefore, the first hypothesis that there would be a difference between learning with the music as opposed to silence was not supported. Similarly, the second hypothesis that there would be a difference between the test with the music as opposed to silence was also not supported.
This finding was consistent with the previous evidence, where students had to remember visually presented digits while listening to instrumental music or silence, which revealed no significant difference in their performance (Alley & Greene, 2008). This may be explained by the fact that similarly to the current experiment, the task involved both visual and auditory stimuli. Therefore the tasks drew on different processes, without one affecting another. This was observed in study by Vredeveldt et al. (2012), where memory performance was affected more when two stimuli presented at the same time involved congruent sensory information as opposed to incongruent. This was also demonstrated by Iwanaga and Ito (2002), where instrumental and vocal music decreased verbal memory recall but not in a visual memory task. However, contrasting results exist, where recall of visually presented digits was disturbed by the background instrumental music (Nittono, 1997).
Additionally, the results showed no interaction between context at learning and context at test. Therefore, the third prediction that there would be an interaction between context at learning and test was not supported. This contrasted with previous research, which found that congruent environmental context at learning and test increased memory performance (Coveney et al, 2013; Godden & Baddeley, 1975; Isarida et al., 2012; Smith, 1982; Thompson et al., 2001; Unsworth et al, 2012). However, the results of the study were consistent with findings by Coveney et at. (2013), who found no difference in medical students’ memory recall between tutorial and operating theatre. Moreover, according to Brinegar, Lehman and Malmberg (2013), the lack of context-dependent effect may have been caused by cued recall. Because the participants in the previous study (Coveney et al., 2013) were familiar with both environments it could have provided cues which helped memory recall.
The current study had a number of limitations, which could have affected the results. Firstly, 2×2 repeated measures ANOVA test revealed small effect size for the context at learning as well as testing and medium effect size for the interaction between the contexts. Provided with a larger population sample, the results may have been different. Secondly, the words used in the task were not matched for the same length, therefore some could have been memorised easier than others, for example ‘shoelace’ and ‘diet’. Finally, because the room in which the study took place was a home setting, the objects around the participant such as flowers, which was one of the words used, could have caused cued recall, which was previously found to eliminate context-dependent effect (Brinegar et al, 2013).
Future research could replicate the study implementing a verbal task as opposed to a visual one to establish the effects of instrumental music on memory recall. The participants should be gathered at the same time to take part in the study, in the room which cannot include objects that can be associated with the words, or alternatively use numbers. It could also include the assessment of the learning style as those who were auditory may have been disturbed by the music less than those who were visual learners. On the other hand, participants who were visual learners could have performed better at visual tasks as opposed to auditory learners, which was previously revealed in another study (Korenman & Peynircioglu, 2007). Also playing action games was found to improve attention, resulting in better allocation of attention simultaneously across several objects (Dye & Bavelier 2010). As discussed earlier, enhanced attention may result in increased encoding of the information and lead to higher memory retrieval (Gillibrand et al., 2011). Controlling the factors discussed could produce more consistent results in the future studies.