Oh Donna if only I had this three weeks ago when analysing a card sort! Wonderful stuff. I look forward to using it.

Although I'm a big fan of running card sorts and learn a great deal each time I do, one of the trickiest parts is quantifying the results. Not any more. Thanks Donna.

Thanks Donna. Just getting into this xls now and I think it is going to be very useful - especially to be able to show the quizzical managers some scientific analysis.

I am not sure how to approach sub categories. One sort group came up with 10 high level categories, but one of those categories has four sub categories, and one has three. Do I use the categories or sub categories in the group column of the Sort worksheet? Using only sub-categories might lose some useful chunking up.

much appreciated
Louisa

Hi Louisa

For analysis, you can do one of two things with subcategories:
- break them into their own little groups, but make a note of the broader group so you don't forget that it was there
- combine them into broader categories and keep a note of the detailed breakdown

The way I make the decision is to look at what the overall pattern is with all the sorters and analyse all at a similar level.

For example, if everyone creates 10 groups with 10 cards, but one group creates 2 main categories with 5 subcategories each, analyse at the sub-category level. If one group creates 10 main groups with 2 sub-groups each, analyse at the broader level.

This isn't just a limitation of my spreadsheet - all statistical analysis needs you to make this sort of decision in order to crunch numbers sensibly.

I'm going to add some notes about this to the instructions when I next review them.

Nick, this spreadsheet has lots of relationships - adding cards means changes need to be made throughout the sheet. I've just sent you the larger spreadsheet, but this too only manages 200 cards. I'll have to make you a spreadsheet that copes with more cards.

In my book, I've made a strong point about thinking about analysis in the planning stage, and making sure you can analyse what you collect ;) It is important to check your analysis tools before you start.

Hi Donna, would you mind sharing the larger spreadsheet with me as well? Thanks!

Donna, I cant think you enough for creating and posting this online. This has been an amazing and eye-opening tool for our card sort. One quick question though - We had 23 participants as opposed to 20. I added a few more tabs to cover their results, added new columns to both summary tabs and changed the formulas, but cant seem to revise the formulas enough pull in the data for tabs Sort21 -Sort23. Any ideas?

Thanks again for sharing!

Donna,

This is great, it seems like it will make my life easier for the card-sort I am doing next week.

I have a question about how you designed the card sort: I see that you put fairly long descriptions on the cards in the first go. How exactly did you elicit the category names from users? Did they create their groups, & assign a group name (original category) of their choosing?

What do you think about the approach where I might provide a label on the front of the card & a brief description of the content on the back?

Thanks.

Hi Donna,

Thank you so much for the spreadsheet - it is invaluable. I wanted to add an extra 15 participants (to have 35 maximum)...is there any way that you can do that easily? I realized that you have to pull forumulas from the different tabs... or do you have a bigger spreadsheet that can accommodate a larger number of participants?

Thanks!

Thanks a lot for this and for your article on card sorting. Really useful!

Hi Sharon

100 people is really too many for a spreadsheet ;) The spreadsheet is great for few enough participants that you can print out the big matricies and look at what they came up with, and you really can't do that with 100 people. Your best bets for that number of participants is:
- split them somehow (demographics, role) and compare results across a couple of groups
- use a statistical tool

Thanks for making this spreadsheet available Donna. What a great tool especially for novices like myself. We just completed a preliminary card sort and your spreadsheet will make explaining the data so much easier. Thanks for all the hard work!

PS --- your user instructions are very well written as well.

Hi Donna,
this method seems like inviting the guests of a restaurant to make the food in the kitchen with the cooks. Isn't the least thing an IA needs "suggestions" from client side on the site structure? I see the comforting rays of mathematics being emitted from this, which can warm a frosty client-IA relationship and give the IA the untouchable aura of a scientist, which then paves the way for the delivery of the actual work: a new site concept that none of the clients could have thought of with more than an initial idea anyway (an idea you can easily collect in a two hour brainstorming meeting). I don't see anything but a soft, mellow middle Information Architecture solution coming out of this that pleases everyone and no one. Unless you throw all data out after everyone got the feeling that their voice counts and will surely embrace the new solution. (The old mediator trick) and then start the actual work. I also can't remember the last project that actually had a budget for something more than a questionaire/brainstorming meeting to gather everyone's opinions on the company, products, their own work and their wishes for the new site. Luxurious! Where did you find the clients that pay for this? The mysterious, undiscovered forests of the Information Architecture market.

This is fantasitc, Donna. Thank you so much for sharing your spreadsheet. What a huge time saver for us all. I wish you every success with your book!

Donna,

Love the spreadsheet. It's an advantage to use with IA, not a detour. I have a question: how do you handle sorters using the same category, but in different ways. For example. You are cateorizing food. Everyone puts meat in a category called meat. But then half the people choose a category called 'yummy stuff', and put in candy and deserts. The other half has a category called 'yummy stuff' and puts in candy and nuts. Seeing this double combo each used, do you rename all of their entries and 'decide for yourself that they got it wrong', and then set your standardized names accordingly. Or do you lump them together and avoid putting too much of your own spin on it? In addition to the two categories above, the first group also has a category called 'nuts', with 'nuts'. The second group has a category called 'desserts', with desserts. So, there are categories that could work for half of each of their 'yummy stuff' entries, but it's different for half the company.

I'd like to react to the commenter above: "this method seems like inviting the guests of a restaurant to make the food in the kitchen with the cooks".

I think this analogy is slightly wrong here. If you say that the guests are creating the food, that's like saying the users are invited to help building the website.

I think a better analogy is that, when a chef in a restaurant prepares a new menu with several new dishes, s/he invites people over and they can taste morsels of food and deciding what flavors they like better. S/he gets a better understanding of what works well and what does not work well. In the end, the final decision of how to make the dish and which dishes to put on the menu, is still up to him/her.

As for budget, I don't think a card sort should be expensive. In the very worst case, it's something you could do with friends or family. And even a small exercise should give you some more insight.

In the end, the user is always right. If they find that a label is unclear or that a certain item does not belong in a category, then that should be catered for, no matter how perfect you think your Information Architecture is.

Thanks - I think I can fix that. I just need to redefine the range in StandardSummary to exclude the card names.

Hi Lilian

Thanks for your comments. I'm probably not going to get time for a while to update the instructions to Word 2007. I did it myself though recently - I think you'll be able to fumble through if you look for the intent of each step.

Thanks a lot for this helpful spreadsheet!

While using it for my card sorting data, I found a little bug I want to let you know, for that it won't keep you as busy as me ;-)

The spreadsheet doesn't allow that 2 different sorters used the same name for one of their categories if these same-named categories are then mapped to 2 different standardised categories. The calculations for this category name then don't work correctly.

Hi Donna, love the spreadsheet, great use of Excel.

I do have a problem with the Unique cards column in the CatsSummary worksheet, though.

For each Standardised category, the number of Unique cards is exactly the same as the number of Total cards in this category, and I know this not to be the case in at least some of the categories.

The ChangeLog says that on 5/11/2010 "Fixed the Unique cards column in CatsSummary". What was the problem that was fixed?

Any thoughts on what could be going on?

cheers, Pete

Hi there,

first thank you for the book and all the knowledge sharing.

I'm desperately trying to make it work on windows 7, I tried with office 2007 and office 2010 but I have run-time errors on both. I'm using the files provided by optimalworkshop.com.
I guess I should use a anterior version of office?

Any idea if somebody made it work with office 2007 or 2010, I can't find any info :-(

Thanks

Yann