38 TESTING 4A 110 Facial Pores - Area ■ Product A ■ Product B 100 4B TABLE
Sidák's multiple comparisons test
Product A D0 vs. D28 D0 vs. D56
90
Product B D0 vs. D28 D0 vs. D56 Test details
80 ** 70 D0 D28
*** D56
Figure 4A and 4B: The importance of rigorous statistical analyses. The choice of the statistical method for analysis of results is crucial for correct interpretations and conclusion. In this example, two-way ANOVA is the correct method, since there are two independent variables (treatment and time) influencing the dependent variable (area of facial pores). It is also important to remark the necessity to include bar errors in the graphical results. To use Paired Student’s t-test for all type of analyses, or including graphs without bar errors, are very common mistakes leading to misinterpretations and misleading conclusions
Strategies to ensure credibility and transparency Beyond the previous defined practices, there are some standard strategies that all the testing laboratories should follow, in order to improve credibility and transparency during clinical trials. With regard to scientific rigour and documentation, employing scientific protocols and standardized methodologies, such as blinded-randomized assessments or placebo- controlled studies, adds credibility and validity to trial outcomes. At the same time, maintaining comprehensive records detailing trial procedures, participant demographics, and observed changes is essential. Independent verification and peer review
might be also a good strategy to increase the relevance of the results. 3 different levels of verification can be established: First, results are obtained after internal ‘informal’ studies, without an external CRO participating in the measurements nor in the analysis of results. Second, the CRO is responsible of the clinical test and the complete final report includes all the raw data, analysis of results, and conclusions are made by an experienced professional. Third, the results obtained in the final report provided by the CRO are published in a scientific journal and peer-reviewed by one or more people with similar competencies. It is widely recognized that the reliability of ‘internal’ results has to be lower than when results are obtained and/or reviewed by external parties such as CRO or scientific journals. Another important strategy involving
marketing is emphasizing on realistic expectations. Educating consumers on realistic outcomes is crucial. Using subjective evaluations to claim high percentages of improvement
PERSONAL CARE January 2025
Product A D0 vs. D28 D0 vs. D56
Product B D0 vs. D28 D0 vs. D56
Sidák's multiple comparisons test
Two-way ANOVA test with Sidák's multiple comparisons between timepoints threshold? Summary Adjusted
Mean Diff
95.00% CI of Diff
15.00 3.928 to 26.07 22.36
11.29 to 33.43
4.076 -6.996 to 15.15 5.000 -6.072 to 16.07 Mean 1
Mean 2
100.0 100.0
100.0 100.00
Mean Diff
85.00 77.64
95.92 95.00
95.00 CI of Diff Below P Value
Yes Yes
No No
15.00 22.36
4.076 5.000
Below
D28 D56
Test details Mean 1
Product A- Product B D0
D28 D56
100.0 85.00 77.64
without having an instrumental objective validation of those results might be bread for today and hunger for tomorrow, since the market has more knowledge today than they had 20 years ago, and they will understand soon that cosmetic topical applications could not give rise to the same ‘percentages’ than clinical surgeries. Avoiding extravagant promises or unrealistic transformations in advertising or promotional material is essential. Moreover, emphasizing that individual responses may vary based on diverse factors helps to manage expectations.
Statistical analysis in cosmetic product evaluations Rigorous statistical analysis stands as an indispensable factor in the realm of cosmetic clinical testing, acting as the keystone for the acquisition of reliable and reproducible data. In the evaluation of cosmetic products’ efficacy and safety, employing robust statistical methodologies is imperative to distil meaningful insights from complex datasets. Statistical analysis not only aids in deciphering the subtle nuances of skin care outcomes but also ensures the precision and accuracy of experimental results. By employing statistical rigour, clinical researchers can effectively quantify the impact of cosmetic interventions, discern trends amidst variability, and ascertain the significance of observed effects, thereby fortifying the credibility and trustworthiness of the findings. This meticulous approach not only elevates
the quality of evidence but also fosters consistency and replicability in outcomes, consequently substantiating the confidence in the conclusions drawn from cosmetic
-32.11 to -2.618 Mean 2
100.0 95.92 95.00
No No Yes
0.000 -10.92 -17.36
** ***
ns ns
Mean Diff SE of Diff
3.949 3.949
3.949 3.949
0.0063 0.0001
0.7828 0.6366 N1
5 5
5 5
P Value
N2
t
DF
5 3.798 16.00 5 5.663 16.00
5 1.032 16.00 5 1.266 16.00
Two-way ANOVA test with Sidák's multiple comparisons between treatments threshold? Summary Adjusted
Product A- Product B D0
0.000 -14.74 to 14.74 -10.92 -25.67 to 3.820 -17.36
ns ns *
Mean Diff SE of Diff
5.746 5.746 5.746
>0.9999 0.1939 0.0176
N1
5 5 5
N2 t DF
5 0.000 24.00 5 1.901 24.00 5 3.022 24.00
clinical trials. The following aspects should be considered for a rigorous and reliable statistical analysis in cosmetic clinical testing:
Data collection and measurement Establishing quantitative metrics for assessing changes in skin characteristics, such as wrinkle depth, skin hydration levels, or pigmentation, is crucial. Using standardized instrumental measurement tools, such as spectrophotometry or image analysis software, instead of subjective clinical observations, ensures consistency in data collection.
Descriptive statistics Utilizing descriptive statistics, such as mean, median, and standard deviation, summarizes the central tendency and variability of data collected from participants before and after product usage, considering the intravariability inherent to human clinical testing. Presenting graphical representations, like box plots or histograms, might be also useful to visually illustrate the distribution of changes observed among participants.
Paired student’s t-tests or Analysis of Variance (ANOVA) The choice of method—paired Student’s t-test, ordinary one-way ANOVA, or two-way ANOVA— relies on the complexity of the experimental setup, the number of variables involved, and the specific research questions being addressed within cosmetic clinical testing. The paired Student’s t-test is suitable when
examining the effects of a single intervention on a single group, making comparisons between two related sets of data points.
www.personalcaremagazine.com
Area (%)
*
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56 |
Page 57 |
Page 58 |
Page 59 |
Page 60 |
Page 61 |
Page 62 |
Page 63 |
Page 64 |
Page 65 |
Page 66 |
Page 67 |
Page 68 |
Page 69 |
Page 70 |
Page 71 |
Page 72 |
Page 73 |
Page 74 |
Page 75 |
Page 76
