Addressing 'Errors' in dataset with null values

[jiexika24]

Hello, when refreshing my dataset, I get thousands of 'Errors.'

These 'Errors' are due to:

1. The cells being empty. My dataset becomes too large and laggy if I replace all the blanks with 'null'.
2. A data type mismatch. For some columns, new data is different from the old data (e.g., was Text, now Whole Number). I don't need to reference these columns with mismatched types but I don't want to remove the columns in case there is a need in the future. 

It seems to give an 'Error' for rows even with data that I need. (Image 1 is PowerQuery; image 2 is my Excel sheet showing that there is data but it will show the whole row as 'Error'):

How can I best address these 'Errors' without removing the data that I do want to keep? Thanks for any pointers!

[jiexika24]

@Anonymous Thanks so much for your input! Some of the survey questions changed over the past 10 years, causing differing data types within certain columns. But your comments made me realize I should just delete those columns in Power Query, since my graphics don't actually use those. Really appreciate it.

[Anonymous · Accepted solution]

I would just delete the "Changed Type" step and then make every column type text. But you really need to deal with the changing data--what's up with that?

--Nate