ACHARON Survey Design Analysis
Santam Chakraborty
22/05/2020
First we will calculate the summary statistics for all question stems.
| Overall (N=22) | |
|---|---|
| set1_dept_colleagues | |
| Sum | 72.000 |
| Mean (SD) | 3.273 (1.279) |
| Median | 3.500 |
| set1_hospital_admin | |
| Sum | 55.000 |
| Mean (SD) | 2.895 (1.449) |
| Median | 3.000 |
| set1_consult_time | |
| Sum | 95.000 |
| Mean (SD) | 4.318 (0.839) |
| Median | 5.000 |
| set1_workload | |
| Sum | 99.000 |
| Mean (SD) | 4.500 (0.740) |
| Median | 5.000 |
| set1_comply | |
| Sum | 97.000 |
| Mean (SD) | 4.619 (0.669) |
| Median | 5.000 |
| set2_happy | |
| Sum | 62.000 |
| Mean (SD) | 2.952 (1.284) |
| Median | 3.000 |
| set2_supported | |
| Sum | 69.000 |
| Mean (SD) | 3.450 (1.234) |
| Median | 4.000 |
| set2_guidance | |
| Sum | 78.000 |
| Mean (SD) | 3.714 (1.231) |
| Median | 4.000 |
| set2_relevant_issues | |
| Sum | 75.000 |
| Mean (SD) | 3.571 (1.248) |
| Median | 4.000 |
| set2_adequate_time | |
| Sum | 85.000 |
| Mean (SD) | 4.048 (1.203) |
| Median | 4.000 |
| set3_pt_wellbeing | |
| Sum | 91.000 |
| Mean (SD) | 4.136 (0.990) |
| Median | 4.000 |
| set3_legal_liability | |
| Sum | 82.000 |
| Mean (SD) | 3.727 (1.316) |
| Median | 4.000 |
| set3_pt_phy_interaction | |
| Sum | 99.000 |
| Mean (SD) | 4.500 (0.913) |
| Median | 5.000 |
| set3_communication | |
| Sum | 102.000 |
| Mean (SD) | 4.636 (0.492) |
| Median | 5.000 |
| set3_rapport | |
| Sum | 88.000 |
| Mean (SD) | 4.000 (1.234) |
| Median | 4.000 |
| set4_seek_care | |
| Sum | 55.000 |
| Mean (SD) | 3.056 (1.392) |
| Median | 3.000 |
| set4_pt_burden | |
| Sum | 81.000 |
| Mean (SD) | 3.682 (1.460) |
| Median | 4.000 |
| set4_pt_satisfaction | |
| Sum | 90.000 |
| Mean (SD) | 4.091 (1.065) |
| Median | 4.000 |
| set4_pt_reflection | |
| Sum | 80.000 |
| Mean (SD) | 3.810 (1.123) |
| Median | 4.000 |
| set4_pt_expectations | |
| Sum | 89.000 |
| Mean (SD) | 4.045 (1.214) |
| Median | 5.000 |
| set4_care_seeking | |
| Sum | 74.000 |
| Mean (SD) | 3.524 (1.365) |
| Median | 3.000 |
| set4_holistic_care | |
| Sum | 83.000 |
| Mean (SD) | 3.952 (1.071) |
| Median | 4.000 |
| set5_decision_making | |
| Sum | 83.000 |
| Mean (SD) | 4.150 (0.933) |
| Median | 4.000 |
| set5_focus_consult | |
| Sum | 82.000 |
| Mean (SD) | 3.905 (1.091) |
| Median | 4.000 |
| set5_burden_hcw | |
| Sum | 93.000 |
| Mean (SD) | 4.227 (0.869) |
| Median | 4.000 |
| set5_coordination | |
| Sum | 75.000 |
| Mean (SD) | 3.750 (1.251) |
| Median | 4.000 |
| set5_outcomes | |
| Sum | 83.000 |
| Mean (SD) | 3.952 (1.284) |
| Median | 4.000 |
| set5_physician_role | |
| Sum | 77.000 |
| Mean (SD) | 3.500 (1.406) |
| Median | 4.000 |
| set6_issue_type | |
| Sum | 92.000 |
| Mean (SD) | 4.381 (0.973) |
| Median | 5.000 |
| set6_cost | |
| Sum | 82.000 |
| Mean (SD) | 3.727 (1.386) |
| Median | 4.000 |
| set6_problem_finding | |
| Sum | 89.000 |
| Mean (SD) | 4.238 (0.995) |
| Median | 4.000 |
| set6_value | |
| Sum | 68.000 |
| Mean (SD) | 3.238 (1.446) |
| Median | 4.000 |
| set6_care_indivisualisation | |
| Sum | 89.000 |
| Mean (SD) | 4.045 (1.133) |
| Median | 4.000 |
Plot the bar charts of the ratings. In this bar chart series, we show the category specific frequency of the the ratings provided by the experts for each option.
Plot the rating provided by each rater for each option.
Check the aggreement between variables. We will first need to transpose the dataframe. The we will be able to apply the functions that calculate aggreement between raters. In this case each option is a subject for which rating is provided by multiple raters.
##
## Call:
## cat_adjusted(.data = agg_data, object = option, rater = name,
## score = rating, weighting = "linear")
##
## Objects = 33
## Raters = 22
## Categories = {1, 2, 3, 4, 5}
## Weighting = linear
##
## Chance-Adjusted Categorical Agreement with Bootstrapped CIs
##
## Observed Expected Adjusted 2.5 % 97.5 %
## alpha 0.694 0.676 0.055 0.018 0.089
## gamma 0.692 0.538 0.333 0.253 0.421
## irsq 0.692 0.674 0.054 0.019 0.085
## kappa 0.692 0.672 0.061 0.028 0.092
## pi 0.692 0.674 0.055 0.018 0.089
## s 0.692 0.600 0.230 0.171 0.297
We now calculate the category specific aggrement indices.
##
## Call:
## cat_specific(.data = agg_data, object = option, rater = name,
## score = rating)
##
## Objects = 33
## Raters = 22
## Categories = {1, 2, 3, 4, 5}
##
## Category-Specific Agreement with Bootstrapped CIs
##
## Estimate 2.5 % 97.5 %
## 1 0.074 0.038 0.100
## 2 0.150 0.112 0.185
## 3 0.123 0.101 0.145
## 4 0.266 0.240 0.288
## 5 0.442 0.388 0.489
We now calculate agreement across categories for each of the 33 questions.
Create a plot of the agreement values for categories in each question
## Warning: Removed 22 rows containing missing values (geom_point).
Make the above plot faceted by the question set.
## Warning: Removed 22 rows containing missing values (geom_point).
Create a dataframe with sum, average, lower bound mean - 2SD and agreement scores across response categories for each option.
Now we take a look at the free text inputs provided by the experts.