Your knack for creative exploration is remarkable. It's like seeing a single visual through different lenses, helping us pinpoint the perfect solution. Your dedication to this process ensures we always achieve excellence.
Hi Bas, thank you for the video - visuals are looking neat and your videos are always very informational! Would love to see more survey based videos! Just to let you know, when you structure your data this way, you miss out on one of the key awesome features that power bi visualizations offer -> drilling down. One of the key insights that survey visuals can provide, is to answer questions such as "How do people that rate Comfort 'Very bad' score on other parameters?". Is comfort an "isolated" issue, or are people who are dissatisfied with comfort also dissatisfied with the facilities, the price, etc. If you think the price was too high, what was it in relation to? If you had an "overall satisfaction" question - which areas affect the overall experience the most, where should you put your efforts to create a better overall experience? With that in mind, if you create more survey answers related videos, I would be very happy! Surveys in PBI is generally not an area with enough content in my opinion!
Thinking it through, I reckon you could create slicers to allow users to view relationships in that way. But if you have a neater solution, I would be all ears! Thank you again!
Cool video, I will grab some tricks from this!! Past week I had been making the same visual result but in another way. In my case the core table has A LOT of this type of likert questions, so I duplicate the core table and then insolated each one, after that appling unpivot operation. Later I throuws Attribute in Y and Value in X and legend. Finally I make a SINGLE calculated measure for each table to make the % appear at DATA LABELS. The DAX I use is this one: example = ROUND(DIVIDE(COUNTROWS('Question_x'), CALCULATE(COUNTROWS('Question_x), ALLSELECTED('Question_x))) *100, 1) & " %" The process ends up as the same but it took less measure creation
Hi Bas, thank you for this. It helped me transform the reporting of performance review distributions by demographics. If you have any ideas on how to demonstrate staff movement between internal departments, I think a lot of HR Analysts would be grateful! Love all your videos
Hi Bas, I would prefer the second. But a likkert scale has 5 (or 7) options. I had a student splitting the middle (3) scale into two parts. Quite some work in Excel, but after seeing your video this is doable in Power BI, for example: 3Right = 0,5 x "3" and 3Left = -0,5 x "3". Another remark: at presenting your survey results you should always present n = 30 (where 30 is the number of respondents), this can be different per question. My ultimate graph survey graph would also take this into account, meaning that the surface of the bar would reflect the number of respondents.
Thanks for sharing, Bas. Very interesting video. I just used the first visual yesterday, and for me this is the best of the three ones you proposed, especially in my case as I had to visualize 10 questions and for that small multiples can become very challenging to understand and focus on/
Excelent class. I prefer the first option, I think is easier to understand, however, it's good show more options to the end users beacuse in the end, they decide what they want to see.
Great content! Thanks! #2 is definitely the least readable. It didn't pass the 5 second rule for anyone on my team when I showed the screenshot. Depending on space I'd go 1 or 3 leaning towards 3.
Great Video! After some consideration i came to the conclusion that i like the second chart option the most, with the shifted bars. Because if i wanted to find i out which areas performs the worst and i looked at option 1, i might be tempted to think its Location, Staff and Money, due to the very bad percentages, are the highest in those categories. However, when i take a look at option 2, it becomes clear that the overall result returns Staff, Money and Comfort as the worst performing, while Location is actually up to a second overall place. The 3 option, i felt also was good to break down each answer respondent by categories, though i felt it lacked the overall picture option 1 and option 2 have
Great video! I do have one question though. What if you have 6 survey different questions: Q1 - Q6 and they all have different responses, how do you work around this?
Excellent video and very insightful I would go for the 3rd option as that easier to read and understand, keep making videos like this as its very helpful....
Thanks for sharing. all three interesting charts. Natural instinct is to go for the top left one but I also liked the third one to compare categories for a given response.
Thank you very much, very rich information that could be used for other applications... Also appreciate very much your way of presenting the illustration, may God bless your heart! I prefer the second one, just wanted to add a constant iline in the middle as it is not understood easily at first
Thanks for making a stats video!! More please!! :-) I work with stats all the time... By the way... at 2mins 12seconds you mentioned another video about unpivoting - I didn't see the link to this :-(
Great video however instead of building measures I just build another table with the answer categories (ie. new table that show 4- good thru 1-very bad) that way you can build a relationship with multiple categories that use the same answers
Hi Bas, I followed your video to build the fourth visual, and it turned out great. However, I am wondering if there's a way to force the titles of the small multiples to always display. For example, in my visual, if no one chose the "very good" option in the survey, the "very good" title disappears. But I would like it to always show. Do you know a way to fix this? By the way, your content is amazing!
Great tutorial! I've been doing the second version for a while, however, I see you have the same issue which is that the diverging bar sort order doesn't match the legend sort order.
@@HowtoPowerBI I think it's more annoying as you can get the same visual as a default in Microsoft Forms. Obviously without any of the other cool functions
That’s actually very funny, on my job I got a servey data for visualization and for now I have already cleaned and prepared the data, but haven’t started doing graphs and… This video has been just uploaded, crazy:)
Nice video 🎉 Just finished project of survey (used the 1st option )now I need to go back right the sketch up table 🤣 to see if the other chart can benefit the end user
Really nice video as always. However, I miss a best practice data model to structure survey data if you have the requirement of being able to cross-filter/cross-highlight questions. I can't see how you can do that with your example here. For instance, you have two visuals. Visual A shows the answers for question A while Visual B shows the answers for question B. Both visuals have a visual-level filter on "Question" (A/B respectively). If you click on a specific answer of one of the visuals, you won't be able to see the combined answers, as both visuals are filtering on "Question" differently. I really lack a way to solve this. Do you have a proper solution for this? Thanks in advance and love your content btw
So how would you calculate say 'Bad' for the category money as a % difference for the total 'Bads' for all categories? Say if Bad for money was 30% and total for all categories was 28% the difference would be 2%? I can't work out how to do that.
Survey projects are not that easy... A lot of data transformation is required... Most of the time surveys capture results in online sheets like google form or smartsheet... I just completed my second survey project, so I know...😅
Your knack for creative exploration is remarkable. It's like seeing a single visual through different lenses, helping us pinpoint the perfect solution. Your dedication to this process ensures we always achieve excellence.
Very good. Every time I watch your videos I learn as much about presenting as I do about Power BI
☺️thanks
How do you always know what I'm working on and provide the solution I'm looking for? This has come at the right time. Thank you!
I’m new to Power Bi. Thank you SO MUCH for this tutorial!!!
Hi Bas, thank you for the video - visuals are looking neat and your videos are always very informational! Would love to see more survey based videos! Just to let you know, when you structure your data this way, you miss out on one of the key awesome features that power bi visualizations offer -> drilling down. One of the key insights that survey visuals can provide, is to answer questions such as "How do people that rate Comfort 'Very bad' score on other parameters?". Is comfort an "isolated" issue, or are people who are dissatisfied with comfort also dissatisfied with the facilities, the price, etc. If you think the price was too high, what was it in relation to? If you had an "overall satisfaction" question - which areas affect the overall experience the most, where should you put your efforts to create a better overall experience? With that in mind, if you create more survey answers related videos, I would be very happy! Surveys in PBI is generally not an area with enough content in my opinion!
Thinking it through, I reckon you could create slicers to allow users to view relationships in that way. But if you have a neater solution, I would be all ears! Thank you again!
Cool video, I will grab some tricks from this!!
Past week I had been making the same visual result but in another way. In my case the core table has A LOT of this type of likert questions, so I duplicate the core table and then insolated each one, after that appling unpivot operation.
Later I throuws Attribute in Y and Value in X and legend.
Finally I make a SINGLE calculated measure for each table to make the % appear at DATA LABELS. The DAX I use is this one:
example =
ROUND(DIVIDE(COUNTROWS('Question_x'),
CALCULATE(COUNTROWS('Question_x),
ALLSELECTED('Question_x)))
*100, 1) & " %"
The process ends up as the same but it took less measure creation
Hi Bas, thank you for this. It helped me transform the reporting of performance review distributions by demographics. If you have any ideas on how to demonstrate staff movement between internal departments, I think a lot of HR Analysts would be grateful! Love all your videos
Hi Bas, I would prefer the second. But a likkert scale has 5 (or 7) options. I had a student splitting the middle (3) scale into two parts. Quite some work in Excel, but after seeing your video this is doable in Power BI, for example: 3Right = 0,5 x "3" and 3Left = -0,5 x "3". Another remark: at presenting your survey results you should always present n = 30 (where 30 is the number of respondents), this can be different per question. My ultimate graph survey graph would also take this into account, meaning that the surface of the bar would reflect the number of respondents.
Thanks for sharing, Bas. Very interesting video. I just used the first visual yesterday, and for me this is the best of the three ones you proposed, especially in my case as I had to visualize 10 questions and for that small multiples can become very challenging to understand and focus on/
Excelent class. I prefer the first option, I think is easier to understand, however, it's good show more options to the end users beacuse in the end, they decide what they want to see.
Great content! Thanks! #2 is definitely the least readable. It didn't pass the 5 second rule for anyone on my team when I showed the screenshot. Depending on space I'd go 1 or 3 leaning towards 3.
you are always doing amazing , thanks a lot for usual support and give us the new tips and tricks in PowerBI , great job brother keep it up.
Another great video. It is always a pleasure to watch Bas. You are both knowledgeable and entertaining.
Another great video, Bas. For me chart no.2 is the best.
Thanks a lot
Great Video! After some consideration i came to the conclusion that i like the second chart option the most, with the shifted bars. Because if i wanted to find i out which areas performs the worst and i looked at option 1, i might be tempted to think its Location, Staff and Money, due to the very bad percentages, are the highest in those categories. However, when i take a look at option 2, it becomes clear that the overall result returns Staff, Money and Comfort as the worst performing, while Location is actually up to a second overall place. The 3 option, i felt also was good to break down each answer respondent by categories, though i felt it lacked the overall picture option 1 and option 2 have
As usual an extremely goo video. Thank you for your efforts.
I would pick the graph #2, very readable to me
Great video! I do have one question though. What if you have 6 survey different questions: Q1 - Q6 and they all have different responses, how do you work around this?
3rd option for me. Much easier to read and I like the formatting, clean and easy on the eye.
😁 it seems that a lot of people have a different preference on this one, .. tricky
Excellent video and very insightful I would go for the 3rd option as that easier to read and understand, keep making videos like this as its very helpful....
Thanks for sharing. all three interesting charts. Natural instinct is to go for the top left one but I also liked the third one to compare categories for a given response.
Thank you very much, very rich information that could be used for other applications...
Also appreciate very much your way of presenting the illustration, may God bless your heart!
I prefer the second one, just wanted to add a constant iline in the middle as it is not understood easily at first
Thanks for making a stats video!! More please!! :-) I work with stats all the time...
By the way... at 2mins 12seconds you mentioned another video about unpivoting - I didn't see the link to this :-(
Great video however instead of building measures I just build another table with the answer categories (ie. new table that show 4- good thru 1-very bad) that way you can build a relationship with multiple categories that use the same answers
This was amazing Basssss!!! Thank you so much!
😁🎉 thx!
Perfect timing …. I am working on similar project 😊
I could feel it 🙃😉
Hi Bas, I followed your video to build the fourth visual, and it turned out great.
However, I am wondering if there's a way to force the titles of the small multiples to always display. For example, in my visual, if no one chose the "very good" option in the survey, the "very good" title disappears. But I would like it to always show. Do you know a way to fix this? By the way, your content is amazing!
I like the first option, as appears, to me, to be clearer. Great video though. thanks Paul
Awesome approach, thanks! I'm going to use the first one in my personal finance dashboard to show the budgetted amount per category
Thanks ☺️ have fun implementing
Great tutorial! I've been doing the second version for a while, however, I see you have the same issue which is that the diverging bar sort order doesn't match the legend sort order.
Yea saw that too after i recorded it lol 😆 , bit annoying but still kinda ok.. i didnt come up yet with an elegant way to fix that
@@HowtoPowerBI I think it's more annoying as you can get the same visual as a default in Microsoft Forms. Obviously without any of the other cool functions
Very nice, it would be nice also to get in the sum of “positive” (good+very good) and “negative” (bad+very bad) in the visualization in some way 🙂
Ah you could actually place those percentages outside of the bars with a dummy that you make invisible
That’s actually very funny, on my job I got a servey data for visualization and for now I have already cleaned and prepared the data, but haven’t started doing graphs and… This video has been just uploaded, crazy:)
I think it is survey time of the year!!! this is the 3d comment that somebody just started working on some survey data lol 😂
I like the conditional formatting measure!
Nice video 🎉
Just finished project of survey (used the 1st option )now I need to go back right the sketch up table 🤣 to see if the other chart can benefit the end user
Oooor you pretend you didnt see the video and save yourself the work 😆
Excellent! Love the CF measure trick.
That fx button makes me happy every time i see it 😆😁
Really nice video as always. However, I miss a best practice data model to structure survey data if you have the requirement of being able to cross-filter/cross-highlight questions. I can't see how you can do that with your example here.
For instance, you have two visuals. Visual A shows the answers for question A while Visual B shows the answers for question B. Both visuals have a visual-level filter on "Question" (A/B respectively). If you click on a specific answer of one of the visuals, you won't be able to see the combined answers, as both visuals are filtering on "Question" differently.
I really lack a way to solve this. Do you have a proper solution for this?
Thanks in advance and love your content btw
hope you create trouble ticket and CRM dashboard with KPI calculations! thank you Mr. Bas!
Once again, very useful video. thank you.
Excellent! I love your sharing so much!
Please create a video on summarized page or an overview page which takes important information from other pages
thx for the idea
Thank you very much, it is very intuitive, 🙏🙏
I would maybe go with small multiples but add 4 rows and 1 column just to have one reference line
So how would you calculate say 'Bad' for the category money as a % difference for the total 'Bads' for all categories? Say if Bad for money was 30% and total for all categories was 28% the difference would be 2%? I can't work out how to do that.
I love it, thanks again Bas !!
Happy to hear that , thx for watching again!!! 😃
In Visual type 2, how do you sort the legend in the correct order? So: Very Bad - Bad - Good - Very Good
Thanks for vid, superb. I'm very enjoy it.
😎👊
I prefer the first one. Others are informative also though.
Good Job Done! Thanks!!!!
😎👊
Awesome, nice video. Thanks for sharing
Thanks for watching! 😃
Very cool, thanks Bas
superb, thank you for sharing
Fantastic, i like all options...kkkk...
hey bro.. i didn't get the dataset ..could you provide link in the chat section..
It is already in the description
amazing
Hi bass, why u multiplied with -1
for the "negative" bars to go to the left
Survey projects are not that easy... A lot of data transformation is required... Most of the time surveys capture results in online sheets like google form or smartsheet... I just completed my second survey project, so I know...😅
i can't able to download the data set
What error do you get