Air Purifier Study
VCH Hygiene
25 September, 2025
Evergreen
Bacteria
Fungi
Spores
Windermere
Bacteria
Fungi
Boxplots with Occupants
Windermere
August 30, 2024
Room 1216 (Resident Room)
Window: open
Door: open
UVC: off
n = 2 for each sample group
Distance from Purifier: 0.3 meters
Fan Speed: 5
Windermere
September 5, 2024
Room 509 (Office)
Window: closed
Door: closed
UVC: off
n = 2 for each sample group
Distance from Purifier: 0.3 meters
Fan Speed: 5
Windermere
September 12, 2024
Room 509 (Office)
Window: open
Door: open
UVC: off
n = 2 for each sample group
Distance from Purifier: 0.3 meters
Fan Speed: 5
Windermere
September 17, 2024
Room 506 (Breakroom)
Window: open
Door: closed
n = 3 for each sample group
Distance from Purifier: 0.3 meters
Fan Speed: 5
Windermere
November 12, 2024
Room 509 (office)
Window: closed
Door: open
UVC: off
n = 2 for each sample group
Distance from Purifier: 0.3 meters
Fan Speed: 5
Windermere
November 13, 2024
Room 509 (office)
Window: closed
Door: open
UVC: off
n = 2 for each sample group
Distance from Purifier: 0.3 meters
Fan Speed: 5
Windermere
November 19, 2024
Room 1015 (resident room)
Window: closed
Door: closed
UVC: on
n = 2 for each sample group
Distance from Purifier: 0.3 meters
Fan Speed: 5
Windermere
November 20, 2024
Room 1015 (resident room)
Window: closed
Door: closed
UVC: on
n = 2 for each sample group
Distance from Purifier: 0.3 meters
Fan Speed: 5
November 25, 2024
Room 1015 (resident room)
Window: closed
Door: closed
UVC: on
n = 2 for each sample group
Distance from Purifier: 0.3 meters
Fan Speed: 5
Windermere
November 27, 2024
Room 1015 (resident room)
Window: closed
Door: closed
UVC: off
n = 2 for each sample group
Distance from Purifier: 0.3 meters
Fan Speed: 5
Windermere
December 3, 2024
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Distance from Purifier: 0.3 meters
Fan Speed: 5
Windermere
December 4, 2024
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Distance from Purifier: 0.3 meters
Fan Speed: 5
Windermere
November 27, December 3 & 4, 2024 Summary
Room 1015 (resident room)
Window: closed
Door: closed
n = 6 for each sample group
Distance From Purifier: 0.3 meter
Fan Speed: 5
Windermere
3 days with UV, 3 days without
Room 1015 (resident room)
Window: closed
Door: closed
Distance from Purifier: 0.3 meter
Fan Speed: 5
Sample: Bacteria
Windermere
3 days with UV, 3 days without
Room 1015 (resident room)
Window: closed
Door: closed
Distance from Purifier: 0.3 meter
Fan Speed: 5
Sample: Fungi
UPAS Data
Purifier on
Purifier off
Windermere
February 25, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group (except bacteria sample group 8 n = 1 because one of the plates had too many colonies to count)
Distance From Purifier: 3.5 meters
Fan Speed: 5
Windermere
February 26, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Distance From Purifier: 3.5 meters
Fan Speed: 5
Windermere
March 4, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Distance From Purifier: 3.5 meters
Fan Speed: 5
Windermere
February 25 & 26, March 4, 2025 Summary
Room 1015 (resident room)
Window: closed
Door: closed
n = 6 for each sample group
Distance From Purifier: 3.5 meters
Fan Speed: 5
Windermere
March 11, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Distance From Purifier: 1.8 meters
Fan Speed: 5
Windermere
March 12, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Distance From Purifier: 1.8 meters
Fan Speed: 5
Windermere
March 18, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Distance From Purifier: 1.8 meters
Fan Speed: 5
Windermere
March 11, 12, 18, 2025 Summary
Room 1015 (resident room)
Window: closed
Door: closed
n = 6 for each sample group
Distance From Purifier: 1.8 meters
Windermere
Distance Comparison
Fan Speed: 5
Different Fan Speed & Distances
Fan Speed = 3
Winderemere
April 22, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Distance from purifier: 3.5 meters
Fan Speed: 3
Winderemere
April 22, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Distance from purifier: 1.8 meters
Fan Speed: 3
Winderemere
April 23, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Distance from purifier: 3.5 meters
Fan Speed: 3
Winderemere
April 23, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Distance from purifier: 1.8 meters
Fan Speed: 3
Winderemere
April 29, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Distance from purifier: 0.3 meters
Fan Speed: 3
Winderemere
May 13, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Distance from purifier: 0.3 meters
Fan Speed: 3
Winderemere
May 14, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Distance from purifier: 3.5 meters
Fan Speed: 3
Winderemere
May 14, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Distance from purifier: 1.8 meters
Fan Speed: 3
Winderemere
June 3, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Distance from purifier: 0.3 meters
Fan Speed: 3
Windermere
Distance From Purifier = 3.5 m
Fan Speed: 3
Windermere
April 22, 23, May 14, 2025 Summary
Room 1015 (resident room)
Window: closed
Door: closed
n = 6 for each sample group
Fan Speed: 3
Distance From Purifier: 3.5 meters
Windermere
Distance From Purifier = 1.8 m
Fan Speed: 3
Windermere
April 22, 23, May 14, 2025 Summary
Room 1015 (resident room)
Window: closed
Door: closed
n = 6 for each sample group
Fan Speed: 3
Distance From Purifier: 1.8 meters
Windermere
Distance From Purifier = 0.3 m
Fan Speed: 3
Windermere
April 29, May 13, June 3 2025 Summary
Room 1015 (resident room)
Window: closed
Door: closed
n = 6 for each sample group
Fan Speed: 3
Distance From Purifier: 0.3 meters
Winderemere
June 4, 2025
Room 1015 (resident room)
Window: open upon arrival, closed for sampling duration
Door: closed
n = 2 for each sample group
Distance from purifier: 3.5 meters
Fan Speed: 1
Winderemere
June 4, 2025
Room 1015 (resident room)
Window: open upon arrival, closed for sampling duration
Door: closed
n = 2 for each sample group
Distance from purifier: 1.8 meters
Fan Speed: 1
Winderemere
June 10, 2025
Room 1015 (resident room)
Window: open upon arrival, closed for sampling durationn
Door: closed
n = 2 for each sample group
Distance from purifier: 0.3 meters
Fan Speed: 1
Winderemere
June 11, 2025
Room 1015 (resident room)
Window: open upon arrival, closed for sampling duration
Door: closed
n = 2 for each sample group
Distance from purifier: 3.5 meters
Fan Speed: 1
Winderemere
June 11, 2025
Room 1015 (resident room)
Window: open upon arrival, closed for sampling duration
Door: closed
n = 2 for each sample group
Distance from purifier: 1.8 meters
Fan Speed: 1
Winderemere
June 17, 2025
Room 1015 (resident room)
Window: open upon arrival, closed for sampling duration
Door: closed
n = 2 for each sample group
Distance from purifier: 0.3 meters
Fan Speed: 1
Winderemere
June 18, 2025
Room 1015 (resident room)
Window: open upon arrival, closed for sampling duration
Door: closed
n = 2 for each sample group
Distance from purifier: 0.3 meters
Fan Speed: 1
Winderemere
June 25, 2025
Room 1015 (resident room)
Window: open upon arrival, closed for sampling duration
Door: closed
n = 2 for each sample group
Distance from purifier: 3.5 meters
Fan Speed: 1
Winderemere
June 25, 2025
Room 1015 (resident room)
Window: open upon arrival, closed for sampling duration
Door: closed
n = 2 for each sample group
Distance from purifier: 1.8 meters
Fan Speed: 1
Windermere
Distance From Purifier = 3.5 m
Fan Speed: 1
Windermere
June 4, 11, 25, 2025 Summary
Room 1015 (resident room)
Window: open upon arrival, closed for sampling duration
Door: closed
n = 6 for each sample group
Fan Speed: 1
Distance From Purifier: 3.5 meters
Windermere
Distance From Purifier = 1.8 m
Fan Speed: 1
Windermere
June 4, 11, 25, 2025 Summary
Room 1015 (resident room)
Window: open upon arrival, closed for sampling duration
Door: closed
n = 6 for each sample group
Fan Speed: 1
Distance From Purifier: 1.8 meters
Windermere
Distance From Purifier = 0.3 m
Fan Speed: 1
Windermere
June 10, 17, 18, 2025 Summary
Room 1015 (resident room)
Window: open upon arrival, closed for sampling duration
Door: closed
n = 6 for each sample group
Fan Speed: 1
Distance From Purifier: 0.3 meters
No Purifier
July 9, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Purifier On First
July 15, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Fan Speed: 5
Distance From Purifier: 0.3 meters
July 16, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Fan Speed: 5
Distance From Purifier: 0.3 meters
Bacteria
July 16, 2025
Room 1015 (resident room)
Window: closed
Door: closed
n = 2 for each sample group
Fan Speed: 5
Distance From Purifier: 0.3 meters
Fungi
July 15, 16 2025 Summary
Room 1015 (resident room)
Window: closed
Door: closed
n = 4 for each sample group
Fan Speed: 5
Distance From Purifier: 0.3 meters
Evergreen
## # A tibble: 1 × 3
## mean_concentration standard_deviation sample_size
## <dbl> <dbl> <int>
## 1 110. 79.4 22## # A tibble: 1 × 3
## mean_concentration standard_deviation sample_size
## <dbl> <dbl> <int>
## 1 89.9 113. 33## # A tibble: 1 × 3
## mean_concentration standard_deviation sample_size
## <dbl> <dbl> <int>
## 1 156. 180. 22## # A tibble: 1 × 3
## mean_concentration standard_deviation sample_size
## <dbl> <dbl> <int>
## 1 114. 128. 33Bacteria
Power Calculation
##
## t test power calculation
##
## n1 = 22
## n2 = 33
## d = 0.2098552
## sig.level = 0.05
## power = 0.116275
## alternative = two.sidedThe effect size with Evergreen bacteria data is 0.210 (small effect size) and with a power of 0.116. The power of 11.6% means that if there is a true difference with an effect size of 0.210, you have an 11.6% chance of detecting it as significant.
Percent Reduction
## [1] 18.52769Data Visualizations
Shapiro-Wilk Normality Test
## # A tibble: 2 × 3
## test shapiro_test shapiro_p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.118
## 2 intervention <htest> 0.000000181Shapiro-Wilk test indicates that the data is not normally distributed.
Data Visualization Log Transformed
## # A tibble: 2 × 3
## test shapiro_test shapiro_p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.00154
## 2 intervention <htest> 0.611Shapiro-Wilk test indicates that the intervention group is normally distributed but the control group is not.
Q-Q Plot
Q-Q Plot indicates that the control group deviates from normal.
Mann-Whitney U Test
##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 441, p-value = 0.1828
## alternative hypothesis: true location shift is not equal to 0This means there is no strong evidence to suggest that there is a significant difference in the bacterial concentrations between the control and intervention groups.
Fungi
Power Calculation
##
## t test power calculation
##
## n1 = 22
## n2 = 33
## d = 0.2685742
## sig.level = 0.05
## power = 0.1599832
## alternative = two.sidedThe effect size with Evergreen fungi data is 0.269 (small effect size) and with a power of 0.160. The power of 16.0% means that if there is a true difference with an effect size of 0.269, you have an 16.0% chance of detecting it as significant.
Percent Reduction
## [1] 26.82707Data Visualization
Shapiro-Wilk Normality Test
## # A tibble: 2 × 3
## test shapiro_test shapiro_p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.0000372
## 2 intervention <htest> 0.00000778Shapiro-Wilk test indicates that the data is not normally distributed.
Data Visualization Log Transformation
Shapiro-Wilk Normality Test
## # A tibble: 2 × 3
## test shapiro_test shapiro_p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.849
## 2 intervention <htest> 0.483Shapiro-Wilk test indicates that the data is normally distributed.
Levene’s Test
## Levene's Test for Homogeneity of Variance (center = median)
## Df F value Pr(>F)
## group 1 0.9721 0.3286
## 53Levene’s Test indicates there is no significant difference in variances between groups.
T Test
##
## Two Sample t-test
##
## data: log_concentration by test
## t = 1.3496, df = 53, p-value = 0.1829
## alternative hypothesis: true difference in means between group control and group intervention is not equal to 0
## 95 percent confidence interval:
## -0.2063013 1.0549712
## sample estimates:
## mean in group control mean in group intervention
## 4.537406 4.113071This means there is no strong evidence to suggest that there is a significant difference in the fungal concentrations between the control and intervention groups.
Windermere
## # A tibble: 1 × 3
## mean_concentration standard_deviation sample_size
## <dbl> <dbl> <int>
## 1 45.1 47.1 324## # A tibble: 1 × 3
## mean_concentration standard_deviation sample_size
## <dbl> <dbl> <int>
## 1 14.6 21.1 494## # A tibble: 1 × 3
## mean_concentration standard_deviation sample_size
## <dbl> <dbl> <int>
## 1 79.2 84.0 323## # A tibble: 1 × 3
## mean_concentration standard_deviation sample_size
## <dbl> <dbl> <int>
## 1 29.6 40.9 488Bacteria
Power Calculation
##
## t test power calculation
##
## n1 = 82
## n2 = 143
## d = 0.9237824
## sig.level = 0.05
## power = 0.9999986
## alternative = two.sidedThe effect size with Windermere bacteria data is 0.924 (large effect size) and with a power of 0.999. The power of 99.9% means that if there is a true difference with an effect size of 0.924, you have an 99.9% chance of detecting it as significant.
Percent Reduction
## [1] 58.62792Data Visualization
Data Visualization Log Transformation
Shapiro-Wilk Normality Test
## # A tibble: 2 × 3
## test shapiro_test shapiro_p_value
## <chr> <list> <dbl>
## 1 control <htest> 7.55e- 7
## 2 intervention <htest> 1.21e-14Shapiro-Wilk test indicates that the data is not normally distributed.
Mann-Whitney U Test
##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 129540, p-value < 2.2e-16
## alternative hypothesis: true location shift is not equal to 0This means there is strong evidence to suggest that there is a significant difference in the bacterial concentrations between the control and intervention groups.
Fungi
Power Calculation
##
## t test power calculation
##
## n1 = 81
## n2 = 136
## d = 0.6567693
## sig.level = 0.05
## power = 0.9965173
## alternative = two.sidedThe effect size with Winderemere fungal data is 0.657 (moderate-to-large effect size) and with a power of 0.997. The power of 99.7% means that if there is a true difference with an effect size of 0.657, you have an 99.7% chance of detecting it as significant.
Percent Reduction
## [1] 57.10764Data Visualization
Data Visualization Log Transformation
Shapiro-Wilk Normality Test
## # A tibble: 2 × 3
## test shapiro_test shapiro_p_value
## <chr> <list> <dbl>
## 1 control <htest> 1.90e- 5
## 2 intervention <htest> 1.89e-12Shapiro-Wilk test indicates that the data is not normally distributed.
Mann-Whitney U Test
##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 119947, p-value < 2.2e-16
## alternative hypothesis: true location shift is not equal to 0This means there is strong evidence to suggest that there is a significant difference in the fungal concentrations between the control and intervention groups.
Windermere UVC
## # A tibble: 2 × 3
## sample test_result p_value
## <chr> <list> <dbl>
## 1 bacteria <htest> 0.4
## 2 fungi <htest> 0.1## # A tibble: 4 × 4
## sample uv median_reduction IQR_reduction
## <chr> <chr> <dbl> <dbl>
## 1 bacteria off 81.5 15.4
## 2 bacteria on 70.5 16.7
## 3 fungi off 84.2 5.59
## 4 fungi on 80.3 14.7Windermere 3.5 m from Purifier, Fan Speed 5
Bacteria
Data Visualization
Data Visualization Log Transformation
Shapiro-Wilk Normality Test
## # A tibble: 2 × 3
## test shapiro_test shapiro_p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.921
## 2 intervention <htest> 0.00306Shapiro-Wilk test indicates that the control group is normally distributed but the intervention group is not.
Q-Q Plot Log Transformed
Q-Q Plot indicates that the control group follows a normal distribution but the intervention group does not.
Mann-Whitney U Test
##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 784, p-value = 5.897e-05
## alternative hypothesis: true location shift is not equal to 0This means there is strong evidence to suggest that there is a significant difference in the bacterial concentrations between the control and intervention groups.
Percent Reduction
## # A tibble: 1 × 7
## mean_concentration_control mean_concentration_interve…¹ standard_deviation_c…²
## <dbl> <dbl> <dbl>
## 1 39.4 10.7 44.9
## # ℹ abbreviated names: ¹mean_concentration_intervention,
## # ²standard_deviation_control
## # ℹ 4 more variables: standard_deviation_intervention <dbl>,
## # sample_size_control <int>, sample_size_intervention <int>,
## # percent_reduction <dbl>Percent reduction = 72.7%
Fungi
Data Visualization
Data Visualization Log Transformation
Shapiro-Wilk Normality Test
## # A tibble: 2 × 3
## test shapiro_test shapiro_p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.0000331
## 2 intervention <htest> 0.0233Shapiro-Wilk test indicates that the data is not normally distributed.
Mann-Whitney U Test
##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 739.5, p-value = 0.001602
## alternative hypothesis: true location shift is not equal to 0This means there is strong evidence to suggest that there is a significant difference in the fungal concentrations between the control and intervention groups.
Percent Reduction
## # A tibble: 1 × 7
## mean_concentration_control mean_concentration_interve…¹ standard_deviation_c…²
## <dbl> <dbl> <dbl>
## 1 29.8 16.9 17.5
## # ℹ abbreviated names: ¹mean_concentration_intervention,
## # ²standard_deviation_control
## # ℹ 4 more variables: standard_deviation_intervention <dbl>,
## # sample_size_control <int>, sample_size_intervention <int>,
## # percent_reduction <dbl>Percent reduction = 43.3%
Windermere 1.8 m from Purifier, Fan Speed 5
Bacteria
Data Visualization
Data Visualization Log Transformation
Shapiro-Wilk Normality Test
## # A tibble: 2 × 3
## test shapiro_test shapiro_p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.572
## 2 intervention <htest> 0.00000397Shapiro-Wilk test indicates that the control group is normally distributed but the intervention group is not.
Q-Q Plot Log Transformed
The Q-Q Plot indicates that the control group follows a normal distribution but the intervention group does not.
Mann-Whitney U Test
##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 892, p-value = 1.44e-07
## alternative hypothesis: true location shift is not equal to 0This means there is strong evidence to suggest that there is a significant difference in the bacterial concentrations between the control and intervention groups.
Percent Reduction
## # A tibble: 1 × 7
## mean_concentration_control mean_concentration_interve…¹ standard_deviation_c…²
## <dbl> <dbl> <dbl>
## 1 30.4 8.10 27.4
## # ℹ abbreviated names: ¹mean_concentration_intervention,
## # ²standard_deviation_control
## # ℹ 4 more variables: standard_deviation_intervention <dbl>,
## # sample_size_control <int>, sample_size_intervention <int>,
## # percent_reduction <dbl>Percent reduction = 73.4%
Fungi
Data Visualization
Data Visualization Log Transformation
Shapiro-Wilk Normality Test
## # A tibble: 2 × 3
## test shapiro_test shapiro_p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.122
## 2 intervention <htest> 0.000265Shapiro-Wilk test indicates that the control group is normally distributed but the intervention group is not.
Q-Q Plot Log Transformed
The Q-Q Plot indicates that the control group follows a normal distribution but the intervention group does not.
Mann-Whitney U Test
##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 963.5, p-value = 5.974e-10
## alternative hypothesis: true location shift is not equal to 0This means there is strong evidence to suggest that there is a significant difference in the fungal concentrations between the control and intervention groups.
Percent Reduction
## # A tibble: 1 × 7
## mean_concentration_control mean_concentration_interve…¹ standard_deviation_c…²
## <dbl> <dbl> <dbl>
## 1 34.8 6.43 17.9
## # ℹ abbreviated names: ¹mean_concentration_intervention,
## # ²standard_deviation_control
## # ℹ 4 more variables: standard_deviation_intervention <dbl>,
## # sample_size_control <int>, sample_size_intervention <int>,
## # percent_reduction <dbl>Percent reduction = 81.5%
Windermere 0.3 m from Purifier Fan, Fan Speed 5 (Room 1015, no UVC)
Bacteria
Data Visualization
Data Visualization Log Transformation
Shapiro-Wilk Normality Test
## # A tibble: 2 × 3
## test shapiro_test shapiro_p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.0954
## 2 intervention <htest> 0.0000366Shapiro-Wilk test indicates that the control group is normally distributed but the intervention group is not.
Mann-Whitney U Test
##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 909.5, p-value = 4.521e-08
## alternative hypothesis: true location shift is not equal to 0This means there is strong evidence to suggest that there is a significant difference in the bacterial concentrations between the control and intervention groups.
Percent Reduction
## # A tibble: 1 × 7
## mean_concentration_control mean_concentration_interve…¹ standard_deviation_c…²
## <dbl> <dbl> <dbl>
## 1 48.8 9.17 38.1
## # ℹ abbreviated names: ¹mean_concentration_intervention,
## # ²standard_deviation_control
## # ℹ 4 more variables: standard_deviation_intervention <dbl>,
## # sample_size_control <int>, sample_size_intervention <int>,
## # percent_reduction <dbl>Percent reduction = 71.2%
Fungi
Data Visualization
Data Visualization Log Transformation
Shapiro-Wilk Normality Test
## # A tibble: 2 × 3
## test shapiro_test shapiro_p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.0863
## 2 intervention <htest> 0.0000000112Shapiro-Wilk test indicates that the control group is normally distributed but the intervention group is not.
The Q-Q Plot indicates that the control group follows a normal distribution but the intervention group does not.
Mann-Whitney U Test
##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 945, p-value = 6.606e-10
## alternative hypothesis: true location shift is not equal to 0This means there is strong evidence to suggest that there is a significant difference in the bacterial concentrations between the control and intervention groups.
Percent Reduction
## # A tibble: 1 × 7
## mean_concentration_control mean_concentration_interve…¹ standard_deviation_c…²
## <dbl> <dbl> <dbl>
## 1 16.0 2.14 9.50
## # ℹ abbreviated names: ¹mean_concentration_intervention,
## # ²standard_deviation_control
## # ℹ 4 more variables: standard_deviation_intervention <dbl>,
## # sample_size_control <int>, sample_size_intervention <int>,
## # percent_reduction <dbl>Percent reduction = -75%
Windermere 3.5 m from Purifier Fan Speed 3 (Room 1015, no UVC)
Bacteria
## # A tibble: 2 × 3
## test shapiro p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.00239
## 2 intervention <htest> 0.000431##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 756, p-value = 0.0006634
## alternative hypothesis: true location shift is not equal to 0## # A tibble: 1 × 3
## control intervention percent_reduction
## <dbl> <dbl> <dbl>
## 1 25.2 11.7 53.7This means there is strong evidence to suggest that there is a significant difference in the bacterial concentrations between the control and intervention groups.
Percent Reduction = 53.7%
Fungi
## # A tibble: 2 × 3
## test shapiro p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.606
## 2 intervention <htest> 0.00789##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 883, p-value = 3.903e-07
## alternative hypothesis: true location shift is not equal to 0## # A tibble: 1 × 3
## control intervention percent_reduction
## <dbl> <dbl> <dbl>
## 1 61.9 17.6 71.5This means there is strong evidence to suggest that there is a significant difference in the bacterial concentrations between the control and intervention groups.
Percent Reduction = 70.9%
Windermere 1.8 m from Purifier Fan Speed 3 (Room 1015, no UVC)
Bacteria
## # A tibble: 2 × 3
## test shapiro p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.0985
## 2 intervention <htest> 0.000104##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 806, p-value = 4.472e-05
## alternative hypothesis: true location shift is not equal to 0## # A tibble: 1 × 3
## control intervention percent_reduction
## <dbl> <dbl> <dbl>
## 1 19.0 7.38 61.1This means there is strong evidence to suggest that there is a significant difference in the bacterial concentrations between the control and intervention groups.
Percent Reduction = 61.1%
Fungi
## # A tibble: 2 × 3
## test shapiro p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.0400
## 2 intervention <htest> 0.000165##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 917.5, p-value = 2.951e-08
## alternative hypothesis: true location shift is not equal to 0## # A tibble: 1 × 3
## control intervention percent_reduction
## <dbl> <dbl> <dbl>
## 1 65 13.5 79.3This means there is strong evidence to suggest that there is a significant difference in the bacterial concentrations between the control and intervention groups.
Percent Reduction = 78.9%
Windermere 0.3 m from Purifier Fan Speed 3 (Room 1015, no UVC)
Bacteria
## # A tibble: 2 × 3
## test shapiro p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.251
## 2 intervention <htest> 0.00000418##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 940, p-value = 3.379e-09
## alternative hypothesis: true location shift is not equal to 0## # A tibble: 1 × 3
## control intervention percent_reduction
## <dbl> <dbl> <dbl>
## 1 82.1 6.67 91.9This means there is strong evidence to suggest that there is a significant difference in the bacterial concentrations between the control and intervention groups.
Percent Reduction = 91.5%
Fungi
## # A tibble: 2 × 3
## test shapiro p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.0493
## 2 intervention <htest> 0.00142##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 868, p-value = 1.033e-06
## alternative hypothesis: true location shift is not equal to 0## # A tibble: 1 × 3
## control intervention percent_reduction
## <dbl> <dbl> <dbl>
## 1 48.8 12.5 74.4This means there is strong evidence to suggest that there is a significant difference in the bacterial concentrations between the control and intervention groups.
Percent Reduction = 74.1%
Windermere 3.5 m from Purifier Fan Speed 1 (Room 1015, no UVC)
Bacteria
## # A tibble: 2 × 3
## test shapiro p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.152
## 2 intervention <htest> 0.323## Levene's Test for Homogeneity of Variance (center = median)
## Df F value Pr(>F)
## group 1 0.2402 0.6257
## 64##
## Two Sample t-test
##
## data: log_concentration by test
## t = 4.0443, df = 64, p-value = 0.0001436
## alternative hypothesis: true difference in means between group control and group intervention is not equal to 0
## 95 percent confidence interval:
## 0.4752739 1.4031411
## sample estimates:
## mean in group control mean in group intervention
## 3.765821 2.826613## # A tibble: 1 × 3
## control intervention percent_reduction
## <dbl> <dbl> <dbl>
## 1 70.2 23.2 66.9This means there is strong evidence to suggest that there is a significant difference in the bacterial concentrations between the control and intervention groups.
Percent Reduction = 66.9%
Fungi
## # A tibble: 2 × 3
## test shapiro p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.114
## 2 intervention <htest> 0.0273##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 804.5, p-value = 6.217e-05
## alternative hypothesis: true location shift is not equal to 0## # A tibble: 1 × 3
## control intervention percent_reduction
## <dbl> <dbl> <dbl>
## 1 177. 59.6 66.3This means there is strong evidence to suggest that there is a significant difference in the bacterial concentrations between the control and intervention groups.
Percent Reduction = 66.3%
Windermere 1.8 m from Purifier Fan Speed 1 (Room 1015, no UVC)
Bacteria
## # A tibble: 2 × 3
## test shapiro p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.0187
## 2 intervention <htest> 0.0106##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 751, p-value = 0.0009537
## alternative hypothesis: true location shift is not equal to 0## # A tibble: 1 × 3
## control intervention percent_reduction
## <dbl> <dbl> <dbl>
## 1 34.6 17.0 50.8This means there is strong evidence to suggest that there is a significant difference in the bacterial concentrations between the control and intervention groups.
Percent Reduction = 50.8%
Fungi
## # A tibble: 2 × 3
## test shapiro p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.00545
## 2 intervention <htest> 0.0750##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 808.5, p-value = 4.974e-05
## alternative hypothesis: true location shift is not equal to 0## # A tibble: 1 × 3
## control intervention percent_reduction
## <dbl> <dbl> <dbl>
## 1 159. 43.5 72.6This means there is strong evidence to suggest that there is a significant difference in the bacterial concentrations between the control and intervention groups.
Percent Reduction = 72.6%
Windermere 0.3 m from Purifier Fan Speed 1 (Room 1015, no UVC)
Bacteria
## # A tibble: 2 × 3
## test shapiro p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.408
## 2 intervention <htest> 0.0201##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 829.5, p-value = 1.241e-05
## alternative hypothesis: true location shift is not equal to 0## # A tibble: 1 × 3
## control intervention percent_reduction
## <dbl> <dbl> <dbl>
## 1 38.1 12.5 67.2This means there is strong evidence to suggest that there is a significant difference in the bacterial concentrations between the control and intervention groups.
Percent Reduction = 67.2%
Fungi
## # A tibble: 2 × 3
## test shapiro p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.107
## 2 intervention <htest> 0.0160##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 923, p-value = 2.24e-08
## alternative hypothesis: true location shift is not equal to 0## # A tibble: 1 × 3
## control intervention percent_reduction
## <dbl> <dbl> <dbl>
## 1 86.7 28.6 67.0This means there is strong evidence to suggest that there is a significant difference in the bacterial concentrations between the control and intervention groups.
Percent Reduction = 67.0%
Windermere Purifier on First (Room 1015, no UVC)
Bacteria
## # A tibble: 2 × 3
## test shapiro p_value
## <chr> <list> <dbl>
## 1 control <htest> 0.117
## 2 intervention <htest> 0.0224##
## Wilcoxon rank sum test with continuity correction
##
## data: concentration by test
## W = 339, p-value = 0.004804
## alternative hypothesis: true location shift is not equal to 0## # A tibble: 1 × 3
## control intervention percent_reduction
## <dbl> <dbl> <dbl>
## 1 29.3 8.44 71.2