01 - survey analysis
Rui
2025-02-17
Data Attrition
## [1] "Original dataset has 368 rows"## Condition Rows_Remaining Rows_Discarded
## 1 2. Do not consent 357 11
## 2 3. Tech_screener failed 333 24
## 3 4. Not in the US 333 0
## 4 5. Not Female 322 11
## 5 6. Use cream less than 3 times a week 273 49
## 6 7. Fail attention check 272 1
## 7 TOTAL 272 96## [1] "Percentage of records kept: 73.91 %"## [1] "Percentage of records discarded: 26.09 %"## [1] "Final filtered dataset (df_cp) dimensions: 272 rows x 74 columns"Those who didn’t do shopping task - thus do not procced to claim question
Survey - product choice
## Column Non_NA_Count Percent_Complete
## age...22 age...22 201 100.00
## use_frequency use_frequency 201 100.00
## edu edu 201 100.00
## race...51 race...51 201 100.00
## income income 201 100.00
## employ employ 201 100.00
## purchase_frequency purchase_frequency 201 100.00
## purchase_amount purchase_amount 201 100.00
## importance importance 201 100.00
## participantId participantId 201 100.00
## Mec_benchmark_1 Mec_benchmark_1 34 16.92
## Mec_study_effet_a_1 Mec_study_effet_a_1 18 8.96
## disclaimer_1_b_1 disclaimer_1_b_1 18 8.96
## Mec_clinical_study_b_1 Mec_clinical_study_b_1 17 8.46
## disclaimer_2_b_1 disclaimer_2_b_1 17 8.46
## disclaimer_3_a_1 disclaimer_3_a_1 17 8.46
## disclaimer_3_b_1 disclaimer_3_b_1 17 8.46
## Mec_clinical_study_a_1 Mec_clinical_study_a_1 16 7.96
## disclaimer_1_a_1 disclaimer_1_a_1 16 7.96
## disclaimer_2_a_1 disclaimer_2_a_1 16 7.96
## Mec_study_effet_b_1 Mec_study_effet_b_1 15 7.46The proportion that choose 1
## [1] "Proportions of Value = 1 with 95% Confidence Intervals:"## Column Proportion Count Total SE CI_Lower CI_Upper
## 1 Mec_benchmark_1 0.5588235 19 34 0.08515380 0.3919221 0.7257250
## 2 Mec_study_effet_b_1 0.4666667 7 15 0.12881224 0.2141947 0.7191387
## 3 Mec_clinical_study_b_1 0.7058824 12 17 0.11051017 0.4892824 0.9224823
## 4 disclaimer_1_b_1 0.8333333 15 18 0.08784105 0.6611649 1.0000000
## 5 disclaimer_2_b_1 0.8235294 14 17 0.09245944 0.6423089 1.0000000
## 6 disclaimer_3_b_1 0.7647059 13 17 0.10287937 0.5630623 0.9663494
## Power analysis to detect difference between:
## - Mec_benchmark_1 (p1 = 0.559)
## - Mec_study_effet_b_1 (p2 = 0.467)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 460.6045
## p1 = 0.5588235
## p2 = 0.4666667
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* group## Power analysis to detect difference between:
## - Mec_benchmark_1 (p1 = 0.559)
## - Mec_clinical_study_b_1 (p2 = 0.706)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 167.5671
## p1 = 0.5588235
## p2 = 0.7058824
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* group## Power analysis to detect difference between:
## - Mec_benchmark_1 (p1 = 0.559)
## - disclaimer_1_b_1 (p2 = 0.833)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 42.87169
## p1 = 0.5588235
## p2 = 0.8333333
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* group## Power analysis to detect difference between:
## - Mec_study_effet_b_1 (p1 = 0.467)
## - Mec_clinical_study_b_1 (p2 = 0.706)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 65.34682
## p1 = 0.4666667
## p2 = 0.7058824
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* group## Power analysis to detect difference between:
## - Mec_clinical_study_b_1 (p1 = 0.706)
## - disclaimer_1_b_1 (p2 = 0.833)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 170.168
## p1 = 0.7058824
## p2 = 0.8333333
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* group## Power analysis to detect difference between:
## - Mec_clinical_study_b_1 (p1 = 0.706)
## - disclaimer_2_b_1 (p2 = 0.824)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 202.8879
## p1 = 0.7058824
## p2 = 0.8235294
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* group## Power analysis to detect difference between:
## - Mec_clinical_study_b_1 (p1 = 0.706)
## - disclaimer_3_b_1 (p2 = 0.765)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 881.8191
## p1 = 0.7058824
## p2 = 0.7647059
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* groupThe proportion that choose 2
## [1] "Proportions of Value = 2 with 95% Confidence Intervals:"## Column Proportion Count Total SE CI_Lower CI_Upper
## 1 Mec_benchmark_1 0.4411765 15 34 0.08515380 0.2742750 0.6080779
## 2 Mec_study_effet_a_1 0.6666667 12 18 0.11111111 0.4488889 0.8844444
## 3 Mec_clinical_study_a_1 0.6875000 11 16 0.11587810 0.4603789 0.9146211
## 4 disclaimer_1_a_1 0.7500000 12 16 0.10825318 0.5378238 0.9621762
## 5 disclaimer_2_a_1 0.6250000 10 16 0.12103073 0.3877798 0.8622202
## 6 disclaimer_3_a_1 0.8235294 14 17 0.09245944 0.6423089 1.0000000
## Power analysis to detect difference between:
## - Mec_benchmark_1 (p1 = 0.441)
## - Mec_study_effet_a_1 (p2 = 0.667)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 75.09565
## p1 = 0.4411765
## p2 = 0.6666667
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* group## Power analysis to detect difference between:
## - Mec_benchmark_1 (p1 = 0.441)
## - Mec_clinical_study_a_1 (p2 = 0.688)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 62.41638
## p1 = 0.4411765
## p2 = 0.6875
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* group## Power analysis to detect difference between:
## - Mec_benchmark_1 (p1 = 0.441)
## - disclaimer_1_a_1 (p2 = 0.750)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 38.44443
## p1 = 0.4411765
## p2 = 0.75
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* group## Power analysis to detect difference between:
## - Mec_study_effet_a_1 (p1 = 0.667)
## - Mec_clinical_study_a_1 (p2 = 0.688)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 7906.567
## p1 = 0.6666667
## p2 = 0.6875
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* group## Power analysis to detect difference between:
## - Mec_clinical_study_a_1 (p1 = 0.688)
## - disclaimer_1_a_1 (p2 = 0.750)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 811.1791
## p1 = 0.6875
## p2 = 0.75
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* group## Power analysis to detect difference between:
## - Mec_clinical_study_a_1 (p1 = 0.688)
## - disclaimer_2_a_1 (p2 = 0.625)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 905.3658
## p1 = 0.6875
## p2 = 0.625
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* group## Power analysis to detect difference between:
## - Mec_clinical_study_a_1 (p1 = 0.688)
## - disclaimer_3_a_1 (p2 = 0.824)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 155.5154
## p1 = 0.6875
## p2 = 0.8235294
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* groupCombine both
## Column Proportion SE CI_Lower CI_Upper Count Total Label
## 1 Benchmark 0.5588235 0.08515380 0.3919221 0.7257250 19 34 Benchmark
## 2 Mec_study_effet 0.5757576 0.08603396 0.4071310 0.7443841 19 33 Study
## 3 Mec_clinical_study 0.6969697 0.08000056 0.5401686 0.8537708 23 33 Clinical Study
## 4 disclaimer_clinical 0.7941176 0.06934458 0.6582023 0.9300330 27 34 Clinical Disclaimer
## 5 disclaimer_degree 0.7272727 0.07752753 0.5753188 0.8792267 24 33 Degree Disclaimer
## 6 disclaimer_consumer 0.7941176 0.06934458 0.6582023 0.9300330 27 34 Consumer Disclaimer
## Power analysis to detect difference between:
## - Benchmark (p1 = 0.559)
## - Mec_study_effet (p2 = 0.576)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 13436.31
## p1 = 0.5588235
## p2 = 0.5757576
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* group## Power analysis to detect difference between:
## - Benchmark (p1 = 0.559)
## - Mec_clinical_study (p2 = 0.697)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 190.9984
## p1 = 0.5588235
## p2 = 0.6969697
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* group## Power analysis to detect difference between:
## - Benchmark (p1 = 0.559)
## - disclaimer_clinical (p2 = 0.794)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 60.86281
## p1 = 0.5588235
## p2 = 0.7941176
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* group## Power analysis to detect difference between:
## - Mec_study_effet (p1 = 0.576)
## - Mec_clinical_study (p2 = 0.697)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 246.0575
## p1 = 0.5757576
## p2 = 0.6969697
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* group## Power analysis to detect difference between:
## - Mec_clinical_study (p1 = 0.697)
## - disclaimer_clinical (p2 = 0.794)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 314.3602
## p1 = 0.6969697
## p2 = 0.7941176
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* group## Power analysis to detect difference between:
## - Mec_clinical_study (p1 = 0.697)
## - disclaimer_degree (p2 = 0.727)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 3503.346
## p1 = 0.6969697
## p2 = 0.7272727
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* group## Power analysis to detect difference between:
## - Mec_clinical_study (p1 = 0.697)
## - disclaimer_consumer (p2 = 0.794)
## Required sample size per group to detect this difference with 80% power (α = 0.05):
##
##
## Two-sample comparison of proportions power calculation
##
## n = 314.3602
## p1 = 0.6969697
## p2 = 0.7941176
## sig.level = 0.05
## power = 0.8
## alternative = two.sided
##
## NOTE: n is number in *each* groupClaim credibility
## [1] "Mean values with 95% confidence intervals:"## Column Mean SD SE CI_Lower CI_Upper n
## 1 claim_benchmark 4.791045 1.471774 0.1038110 4.587575 4.994514 201
## 2 Original claim 4.750000 1.414214 0.2500000 4.260000 5.240000 32
## 3 Consumer perception 4.878788 1.268798 0.2208694 4.445884 5.311692 33
## 4 High price 4.764706 1.102582 0.1890912 4.394087 5.135325 34
## 5 Low price 5.029412 1.058452 0.1815230 4.673627 5.385197 34
## 6 Smaller participants 4.705882 1.291685 0.2215221 4.271699 5.140066 34
## 7 Big efficacy number 5.235294 1.518748 0.2604631 4.724786 5.745802 34
Consensus and magnitude confusion
## [1] "Mean values with 95% confidence intervals for confidence degree variables:"## Column Mean SD SE CI_Lower CI_Upper n
## 1 Con-degree-high 2.949495 1.053408 0.10587147 2.741987 3.157003 99
## 2 Con-degree-low 2.598039 0.925543 0.09164241 2.418420 2.777658 102
## Basic statistics:## Con-degree-high: Mean = 2.949495 , SD = 1.053408 , n = 99## Con-degree-low: Mean = 2.598039 , SD = 0.925543 , n = 102## Difference (high - low): 0.3514557## One-sided t-test (Con-degree-high > Con-degree-low):##
## Welch Two Sample t-test
##
## data: high_confidence and low_confidence
## t = 2.5099, df = 194.13, p-value = 0.006446
## alternative hypothesis: true difference in means is greater than 0
## 95 percent confidence interval:
## 0.1200301 Inf
## sample estimates:
## mean of x mean of y
## 2.949495 2.598039##
## Hypothesis Test Summary:## -------------------------## Null Hypothesis: Mean of Con-degree-high is less than or equal to mean of Con-degree-low## Alternative Hypothesis: Mean of Con-degree-high is greater than mean of Con-degree-low## Test statistic (t): 2.51## Degrees of freedom: 194.127## p-value: 0.006446## Decision: Reject the null hypothesis at the 0.05 significance level.
## Conclusion: There is sufficient evidence to conclude that the mean of Con-degree-high
## is significantly greater than the mean of Con-degree-low.##
## Effect size (Cohen's d): 0.355## Interpretation: Medium effect size