Help needed to conduct a statistical test to measure statistically significant increase/decrease while comparing one Provider Overall Experience value to Overall Provider Overall Experience value

🔍 Calling all Tableau enthusiasts and wizards! 🚀

I'm currently tackling an urgent client challenge and could really use your expertise! 📊💡

Here's the scoop: I need to analyze if the top 3 ratings (8, 9, 10) for provider 'B' are statistically significant increase or decrease as compared to the combined top 3 ratings (8, 9, 10) of all providers in our survey results dataset. 📈✨ This analysis needs to extend to other providers too.

I've attached both the sample data source and the Tableau file to this question for your reference. 📎

Would anyone be able to lend a hand in crafting the right analysis and calculations within Tableau? I highly appreciate the community and help in advance.🤝💻 #Tableau ⌗DataAnalysis ⌗UrgentHelpNeeded

Here's some context about the dataset and the calculations I've attempted:

1. The dataset has providers represented in a column named 'Provider', and each row contains a unique identifier 'RespondentId' along with their corresponding 'Overall Experience'. However, there are duplicate values in the 'RespondentId' field.
2. I've created calculations for the top 3 ratings (8, 9, 10) for provider 'B' and ensured unique respondent IDs.
3. I'm encountering errors while attempting the t-test calculations, considering the dataset's background context.

I've provided the underlying calculations for mean, standard deviation, and sample size for both provider 'B' and all providers (including 'B').

Could anyone who has experience with Tableau help me ensure that my t-test calculations incorporate this background context appropriately?

1. Calculation when rating is greater than 7

OET3B =>

(IF [Overall Experience] > 7 THEN 1

ELSEIF [Overall Experience] <=7 THEN 0

END).

2. Since the dataset has duplicate respondent ids we need unique respondent ids. So, the following calculation is added in the sheet

'OE T3B - Unique ID'=>

"{ FIXED [RespondentId] : AVG([OE T3B]) }"

3. We need to make sure that the t test calculations take this background context and then create appropriate calculations.

(ABS([Mean for provider 'B'] - [Mean for all providers (including 'B')]) /

SQRT(([Standard deviation for provider 'B']^2 / [Sample size for provider 'B']) + ([Standard deviation for all providers (including 'B')]^2 / [Sample size for all providers (including 'B')])))

---------------------------------------------------------------------------------------------------------------

Underlying calculations

// Mean for provider 'B'

{FIXED : AVG(IF [Provider] = 'B' THEN [OE T3B - Unique ID] END)}

// Mean for all providers (including 'B')

{FIXED : AVG([OE T3B - Unique ID])}

// Standard deviation for provider 'B'

{FIXED : STDEV(IF [Provider] = 'B' THEN [OE T3B - Unique ID] END)}

// Standard deviation for all providers (including 'B')

{FIXED : STDEV([OE T3B - Unique ID])}

// Sample size for provider 'B'

{FIXED : COUNT(IF [Provider] = 'B' THEN [Respondent ID] END)}

// Sample size for all providers (including 'B')

{FIXED : COUNT([Respondent ID])}