The Rey-Osterrieth complex figure test (ROCFT) is one of the most commonly used neuropsychological tests for assessing visuospatial construction ability and visual memory [1] . Lu et al. described recognition scores of the ROCFT [2] . This test involves copying a complex geometric figure and then reproducing it from memory, either immediately or after a delay, in patients with brain injury. It is widely used in a clinical setting.

Studies using the ROCFT have revealed visual memory disturbance in individuals with schizophrenia [3] . Similarly, individuals with Alzheimer’s disease (AD) and Korsakoff’s syndrome have shown poorer copy and recognition on the ROCFT than controls [4] . Cuyas, Verdejo-Garcia, Fanundo, et al. attempted to clarify the association between 3, 4-methylenedioxymethamphe-tamine (MDMA) cumulative use and cognitive dysfunction [5] . They found that lifetime cumulative MDMA use was significantly associated with poorer performance on visuospatial memory by ROCFT.

In addition, Theppitak, Lai, Izumi, et al. reported that visual recognition and memory performance for elderly persons were improved by extending the encoding time and performing repeated test trails [6] . Moreover, Yamashita (2008) reported that repeated administration of the ROCFT continued significant practice effects for 12 months. In conclusion, a more careful attitude was required for repeated administration of the ROCFT and interpretation of the results [7] .

A few reports have described that the effectiveness of strategies to promote immediate visual memory and delayed recall is measurable using the ROCF test. The results of ROCFT have been reported to vary depending on the memorization method or environmental conditions. For example, the effects of outer speech in a quiet environment were confirmed by Lezak (2004), with “outer speech” referring to a method of expressing one’s own behavior by vocalizing it [8] . We hypothesized that the performance of visual memory is reduced by blocking outer speech.

Two experiments were conducted using the ROCF test. In Experiment 1, the figure was copied under 3 conditions prior to its reproduction: copying it as usual; using outer speech; and holding a mouthpiece in the mouth to prevent such speech. The figure was reproduced and assessed in a quiet environment 3 minutes and 3 days after copying.

In Experiment 2, the test was also conducted under the same conditions, but with noise. As the clinical rehabilitation setting tends to involve various types of noise, the necessity of examining the influence of noise on these conditions was considered.

Thirty (6 males and 24 females with a mean age of 21.20 ± 1.80 years) students belonging to the study university, Kanazawa University, were examined with their consent. All the students were high level university students. They have no mental disorders. They were divided into 3 groups: copying, outer speech, and mouthpiece groups. Each group was made up of 10 members, at the same sex ratio.

This study was conducted with the approval of the Medical Ethics Committee of Kanazawa University.

The test was conducted on a face-to-face basis in a quiet, private room without noise (quiet environment) in all cases. The ROCF figure was reproduced 3 minutes and 3 days after copying it under 3 conditions for memorization: copying as usual; using outer speech; and holding a mouthpiece in the mouth. The outer speech group copied the figure while vocalizing the process of copying, while the mouthpiece group copied it with a mouthpiece held in their mouths.

After copying the figure, all of the subjects were instructed to participate in a discussion to prevent them from rehearsing the reproduction of the figure. They were also instructed to reproduce it after 3 days without notification. The reproduced figure was scored by the examiner based on appropriate criteria. The maximum full score was 36.

A questionnaire survey was conducted after the completion of the task to examine the subjects’ reflection on each condition through free recall.

To compare time-dependent changes in the score under each condition, a 2-way factorial analysis of variance was performed, with the condition as an unpaired factor and the time as a paired factor. This was followed by multiple comparisons using Tukey’s method.

Table 1 shows the mean scores and standard deviations of each group at 3 minutes and 3 days after copying. The copying and outer speech groups showed the highest scores at the former and the latter, respectively. The mouthpiece group showed the lowest score at both time points. Furthermore, the score at the latter was lower than that at the former under all conditions. There are significant dif-