For aggression, the score of the experimental group decreased from 44.12±6.66 to 43.48±8.11(
p=.754), and that of the control group also decreased from 45.77±6.91 to 44.75±7.97(
p=.584), but there was no statistical significance. This was not consistent with the result by
Choi (2016) that applied the horticultural activity program based on validation therapy to elderly patients in nursing homes, measured their anger with Vibraimage, and discovered that the aggression score has decreased. It was also not consistent with
Sung and Lee (2016) that applied group validation therapy to patients with mild dementia at a day care center and used the survey method that did not use the same tool for measurement but reported that aggression has decreased. The study by
Choi (2016) included patients with dementia at one nursing home, but this study is different in that patients with dementia are excluded from the subjects. While homogeneity was secured in the scores of effect measurement variables and age of the subjects recruited from the two facilities, exogenous variables according to the internal/external environment of the facility and other characteristics were not controlled, which may have affected the results. Therefore, it is necessary to further elaborate the research design and replicate the experiment. Stress increased the experimental group from 23.63±6.88 to 26.66±7.64 (
p=.095) and in the control group from 23.05±8.04 to 23.60±6.75(
p=.752), but there was no statistical significance. Tension showed a statistically significant increase in the experimental group from 19.75±5.85 to 38.39±18.91(
p<.001), and also in the control group from 19.91±7.83 to 33.82±20.44 (
p=.001). This supports the report that tension increases in senescence due to physical aging and increased rigidity (
Korean Neuros Pyschiatric Associaton, 2017). The horticultural activity program based on validation therapy could not alleviate tension, and in the fact that the participants feel anxious when their tension is high, this result did not support
Koh (2017) that used the survey method instead of the same assessment tool but reported that the horticultural activity of raising companion plants was effective for anxiety, which is a sub-factor of mental health, for the elderly living alone. It is necessary to later verify the effects of a program adjusted so that it does not increase tension about new activities by significantly lowering the intensity of activity for the elderly patients in long-term care nursing homes while using the same horticultural activities. Suspect showed a statistically significant increase in the experimental group from 29.31±3.77 to 36.19±6.10(
p<.001), and also in the control group from 29.90±3.78 to 34.06±6.43 (
p=.006). This was different from the results of
Sung and Lee (2016) reporting that when
Schrijnemaekers et al. (2002) applied group validation therapy, it did not have a statistically significant effect on controlling problem behavior of patients with dementia, but in clinical observation there was a positive change in mental and behavioral symptoms and had motivation to control negative behavior, with which this study combined horticultural activities. This may be because both groups could not be relieved as they become more suspicious. Thus, it is necessary to conduct research on the effects when extending the short-term 10-session program to a long-term program for elderly patients in nursing homes aged 70 and above on average to secure the time to build trust with the subjects.
Balance decreased in the experimental group from 70.00±8.34 to 66.38±8.89(
p=.148), and also in the control group from 67.00±11.25 to 63.07±11.76(
p=.124), but there was no statistical significance. This was not consistent with
Hwang and Yeo (2001) that did not use the same assessment tool but used the survey-type emotional balance scale and reported that the emotional balance scores of 8 elders with dementia increased after horticultural therapy. Since the elderly patients in this study were not diagnosed with dementia, they may have shown different responses depending on their cognitive functions toward validation therapy that acknowledges everything about their current state without asking why. Accordingly, it is necessary to measure the effects later in consideration of cognitive functions. Charm decreased in the experimental group from 78.29±5.86 to 75.67±7.62(
p=.087), and also in the control group to 76.96±74.49(
p=.253), but there was no statistical significance. Energy decreased in the experimental group from 34.40±9.93 to 32.52±10.22(
p=.486), and also in the control group from 39.82±11.41 to 36.87±9.81(
p=.184), but there was no statistical significance. This result was not consistent with
Choi (2016) that the horticultural therapy program based on validation therapy significantly increased energy of the elderly in facilities. This is in line with the result that did not support the result about aggression. This study is different in that it excluded patients with dementia, and thus it is necessary to repeatedly carry out this program considering cognitive functions in the future. Self-regulation decreased in the experimental group from 73.41±6.54 to 70.99±6.89 (
p=.158), and also in the control group from 71.64±6.40 to 68.53±6.71(
p=.099), but there was no statistical significance. This was not consistent with
Kim (2017) reporting a significant difference in self-regulation from the brain function test through electrocorticography on the elderly with mild cognitive impairment by combining forest therapy with horticultural therapy and play therapy as an integrated nature-friendly program. Inhibition decreased in the experimental group from 18.52±2.82 to 15.98±4.12(
p=.012), and also in the control group from 19.47±17.03 to 17.03±4.71 (
p=.028) with a statistically significant difference. Such a significant change is identified because both groups are receiving care from a facility and thus their emotional inhibition has decreased. The score decreased more in the experimental group than the control group, but there is a limitation in that the effect of the program was not verified by checking the amount of decrease according to time because by making an assessment later. Neuroticism showed a statistically significant decrease in the experimental group from 31.64±10.94 to 22.87±13.79(
p=.003), but the decrease in the control group from 37.17±13.93 to 30.87±19.75(
p=.099) was not statistically significant. This result proved that the program is effective in the neurotic psychophysiological state, and supported the results of
Song (2012), who applied the horticultural therapy program once a week for total 12 sessions to promote mental health and improve physical functions of 19 elders aged 65 and above living in long-term care nursing homes. Although there was a difference in the measurement tool, the results showed that mental health condition improved in both the experimental group and control group, but the improvement was greater in the experimental group, and the program was also found effective in the sub-categories such as somatization, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, paranoia and psychosis. Moreover, the positive domain decreased without statistical significance in the experimental group from 63.98±5.45 to 61.39±6.02(
p=.056), but it decreased with statistical significance in the control group from 63.89±5.09 to 60.74±5.48(
p=.036). Considering that there is no statistically significant difference even though the positive function decreased in the experimental group unlike the control group, the effect of this program seems to be maintaining the positive function. This supports
Pyun (2019) that did not use the same assessment tool but reported in a study on healthy elderly aged 65 and above that horticultural activities had a positive impact on psychophysiological effects, and
Yoon and Sung (2017) reporting that the horticultural therapy program on the elderly with mild dementia reduced depression and resulted in positive emotional state. The negative domain increased with statistical significance in the experimental group from 29.38±3.75 to 36.18±6.21(
p<.001), and in the control group from 29.69±3.74 to 34.06±34.06±6.61(
p=.004). This proved that this program was not effective in reducing negative mental functions. The physiological domain decreased with statistical significance in the experimental group from 25.08±6.27 to 19.42±8.80(
p=.003), and also decreased in the control group from 28.82±8.45 to 23.95±12.01(
p=.061), but with no statistical significance. This suggests that the physiological domain reflects self-inhibition and neuroticism and showed a significant decrease in neurotic state, and thus this program had a positive effect on mental functions.