No Date table?
1 Like
Lol I totally forgot about that, since I haven’t started creating measures yet! Thanks!
1 Like
Thanks - mine is nearly identical (plus the extended date table and merging status tables), now that I’m linking Regions on the proper field. 
- Brian
5 days into the challenge and have only set up relationship tables so far 
atleast you have started 
1 Like
hello sunip, me also i couldn’t find any table to join with the Admin ID column
Also, what about client satisfaction column:
- N/A -
- SAT -
- NSA -
Hi All,
Hope you are keeping well.
Apologies it was a bank holiday in the UK so couldn’t reply sooner.
I think from reviewing the posts most of you have manged to create the correct relationship.
Some of you are asking about Administrator Id and which dimension this links to you can ignore this as the dimension is not in the data set.
With regards to the statuses and the history, the aim of the model was to capture when the status of a claim was changed i.e how long was a claim sat in in progress before being resolved or closed.
Every claim does not need to go through every status, the data has been anonymised.
Thanks,
Haroon
1 Like
I interpreted the data as:
SAT - satisfied
NSA - not satisfied
N/A - client satisfaction data not available
- Brian
6 Likes
Sorry I missed this.
Thanks Brian your interpretation is correct.
Regards,
Haroon
2 Likes
is it me or some else spotted this phenomenon
Do we need to verify the validity of tables (Correct or incorrect information) ,
for example in customer table the customer age is 1 year old and married ? or customer age is under 18 and the marital status is either Married or divorced? hmm… correct me if I am wrong thank you.
@Melissa and I were discussing this yesterday. Not sure what is more surprising – the number of under 18 marriages and divorces in the dataset, or the number of elementary schoolers fiscally savvy enough to be buying insurance…
- Brian
7 Likes
Well I kind of make this logic (having previous experience with Electronic Health Record) " GIGO , the inference I made is the record is faulty and need to be reviewed so this is my logic at customer level within that table :
IF CORRECT INFO Given= **
SWITCH(TRUE(), Customers_Tbl[AGE]>=18 && (Customers_Tbl[MaritalStatus]=“Single” || Customers_Tbl[MaritalStatus]=“Married” || Customers_Tbl[MaritalStatus]=“Divorced” || Customers_Tbl[MaritalStatus]=“Deceased”),“Correct”,
Customers_Tbl[ParticipantType]=“Company”,“Correct”,“Incorrect”)
I wouldn’t worry about the validity of data otherwise you would find it difficult to complete the report. In real life it’s unlikely to have these types of error/mistakes. If it does then brace for for many sacking of employees
As it is mentioned that company is forward thinking and Elon Musk is working on Neuralink technology so now everything is possible
Hi All,
Please find attached the link to my dashboard:
I really would love to hear your feedback, all tips, tricks, remarks etc are welcome, as I really want to learn more about working with data and Power BI.
I’m looking forward to see some of the other reports!
Kind regards, Marieke
4 Likes
Great entry. Looks well structured, and has a great flow. See a little error on one of the pages:
Thanks
Jarrett
Nice
Just make sure to add images of your report in the thread so that everyone can view it as they scroll down.
Sam
I totally agree with you in terms of not letting the quirks in the Challenge data hinder your report development, BUT I do think it’s a best practice that as soon as you build your data model to first use it to assess the reliability and validity of the data you’ll be analyzing. Even in many real-world cases, doing this assessment first will save a lot of time in highlighting data that are not of sufficient quality to support further analysis for decision-making.
So far, much of my effort on this challenge has been focused on this issue, and I think I’ve got some interesting techniques that I’m eager to share and discuss.
- Brian
1 Like