This document provides an overview of using inductive content analysis to structure information without user research. It describes analyzing viewer comments on meditation videos by tagging them with keywords, grouping related tags into themes, and naming the themes. These same techniques can be applied to an existing website's content inventory by tagging pages, identifying related tags, and naming the resulting categories. The process involves tagging entries individually, building categories from related tags, and potentially reviewing the analysis with others.
1 of 35
Download to read offline
More Related Content
Guerrilla IA - content analysis when you can't do a card sort
1. Guerrilla IA Techniques
Content Analysis
When You Cant Do a Card Sort
Elizabeth Buie
Northumbria University
Newcastle upon Tyne
#wiad
@ebuie
9. Background
Research paper:
analysis of viewer
comments on
YouTube
meditation videos
Goal: Explore
how people
experience these
videos
Objective: Identify
themes in sample
of comments
9
12. Background: 1354 comments (~14 each)
Reading for my exams, and your videos are perfect as background noise, keeps me from
getting too distracted. Thank you.
I happen to 鍖nd clear skies and bright sunshine depressing, and I have felt this way since early
childhood. The sound of rainfall lifts my spirit considerably when the weather is hopelessly
stuck in sun mode. [username], you're a humanitarian.
Everything has a vibration. The Earth, Sun, Moon, Your Physical Body. Thoughts are electrical
current at which is a wave form or "vibration". It is proven that a speci鍖c stages of thought or
relaxation your mind produces a speci鍖c frequency. Rather then open your mouth with
thoughtless speech to text, open your mind to educate yourself to something that has been in
existence since before time itself.
His voice just doesn't work for me a good message though
I'm trying to 鍖nish my French assignment to this .. I don't think it's going so well .....
shit, I'm trying to awake from my lucid dream
This helped me a lot when i was heavily depressed, it still does.
Atheism was never a religion, not even in the Soviet Union. The belief in God was not
permitted in the soviet union because the totalitarian state wanted the people to put their faith
in them, and not a God. Atheism is not a static movement with a speci鍖c ideology and set of
laws. It only means one does not believe in God. It's like saying "I like pie.
WHOOOAAA ! ! ! THAT WAS INSANELY AWESOME ! ! !
16. Inductive Content Analysis
Tag
each item with its major keyword(s)
Group
related keywords
Name
groups (themes)
It is IA.
(Just without the user studies.)
17. Tag with keywords
Damn this one made my eyes twitch so much
I felt the top of my head tingling
my body was in sleep paralysis mode, then my whole body
started vibrating
.. sent shivers down my spine
got me goosebumps
Felt my spine tingle during it! Amazing
Beautiful music my soul and body started dancing by the rhythm of
this music, excellent
So very beautiful I felt so much energy thank you blessings
Oneness blessing , thank you
Amazing sound. Feels like a brain massage.
WHOOOAAA ! ! ! THAT WAS INSANELY AWESOME ! ! !
18. Tag with keywords
Damn this one made my eyes twitch so much
I felt the top of my head tingling
my body was in sleep paralysis mode, then my whole body
started vibrating
.. sent shivers down my spine
got me goosebumps
Felt my spine tingle during it! Amazing
Beautiful music my soul and body started dancing by the rhythm of
this music, excellent
So very beautiful I felt so much energy thank you blessings
Oneness blessing , thank you
Amazing sound. Feels like a brain massage.
WHOOOAAA ! ! ! THAT WAS INSANELY AWESOME ! ! !
19. Group related keywords
twitch
thank you
amazing
music
tingle/tingling
blessing
beautiful
sound
sleep paralysis
blessings
excellent
vibrating
shivers
goosebumps
dancing
energy
brain massage
body
spine
head
eyes
awesome
20. Name the themes
Thanks
Overall
Features
twitch
thank you
amazing
music
tingle/tingling
blessing
beautiful
sound
sleep paralysis
blessings
excellent
Physical
vibrating
shivers
goosebumps
dancing
energy
brain massage
body
spine
head
eyes
awesome
21. Final comment structure
The Video or its
Creator
Subjective
Experience
Impressions
Relaxation/sleep
Thanks/
blessings
Health/wellbeing
Speci鍖c features
Question
Physical effects
and sensations
Focus/activities
Being
elsewhere
Drugs/alcohol/
herbs
Anger/fear
Humour
Response
Others
to
Advice,
explanation and
support
Re鍖ection
22. The paper
Meditations on YouTube
Elizabeth Buie and Mark Blythe
@ebuie
@markblythe
Northumbria University
Department of Media and Communication Design
24. Start with the content inventory*
*Assumes redoing an existing site, but the technique should work for a new site as well.
25. Prepare the data
Duplicate the content inventory
spreadsheet
Eliminate menu pages that
have no real content
Un-indent; move all descriptors
to Column A
Sort the list alphabetically (not shown)
26. Tag the entries (pages)
Tag every entry with its
pages major keyword(s)
If a page needs more than
one tag, duplicate its entry
Ensure that each entry
has a single tag
27. Build the categories
Identify
and group
related tags
Experiment
with
the groupings
Name
the
categories
s
IA
n
k
w
o
h
w
o
th
w
is
s.
rk
o
34. References
P. Mayring, Qualitative Content Analysis:
www.qualitative-research.net/index.php/fqs/article/view/
1089/2386
S. Elo & H. Kyng辰s, The Qualitative Content Analysis
Process:
www.ncbi.nlm.nih.gov/pubmed/18352969
E. Buie & M. Blythe, Meditations on YouTube:
Paper: leisurelyseekingdoctorate.鍖les.wordpress.com/2013/05/
p41-buie.pdf
際際滷s: www.slideshare.net/ebuie/meditations-on-you-tube
#13: I harvested comments from the selected videos, aiming for a rich sample rather than a representative one.
#14: I did a word cloud of the 1000 most frequent words in the comments, after spell checking the set. Im not sure this cloud shows all 1000, but you can see there are a lot.
#17: Here are the steps in inductive content analysis. Im going to show you what I did with the viewer comments and then explain how I think the technique can apply to IA projects.
#18: Here is a tiny sample of comments. Its not a random sample, but shows some comments that have related keywords.
#19: Here is the sample, with keywords underlined.
(I also flagged whether each was positive, negative or neutral/ambiguous, but that doesnt apply to IA.)
#20: Grouping the keywords allowed me to identify themes in the comments.
#23: I presented the paper at the 2013 Designing Pleasurable Products and Interfaces conference, held in Newcastle upon Tyne last September. (This conference really needs a new name. :-)
Buddha statue photograph by Daniela Hartmann, retrieved from Flickr and used by permission under Creative Commons licence
Laptop photograph by Elizabeth Buie
Video image on laptop from Buddhist Chant - Shingon
#24: Here are some thoughts about how this technique might be useful to an IA effort.
#25: This is a partial content inventory of the bristol.gov.uk website, taken from the site map.
#26: After youve prepared the sanitised spreadsheet, try sorting it alphabetically by page name. This will reduce the chances that the current site structure will unduly influence your tagging.
#27: One beauty of the Web is that content can sit in more than one place. Unlike most card sort methods Ive seen, this technique allows for that.
When you give each entry a single tag, you can treat the tags independently and group and sort them in ways that make sense.
Tag image by Nemo, obtained from Pixabay, public domain - http://pixabay.com/en/blue-sticker-tag-tags-cloths-36821/
#28: As IAs, we already know how to do this part. We can use spreadsheets, sticky notes, card-sorting software whatever works for us.
Photo credit: Elizabeth Buie. Setting: Washington DC.
#29: Youre probably thinking that this sounds essentially like a one-person card sort.
Tagging is what makes it different.
Photo credit: Mike Pennington, http://www.geograph.org.uk/photo/1011825. Used under Creative Commons.
#30: This method involves unpicking the content quite a bit more than we usually do. We see more of the details and become less influenced by the existing structure.
Photo credit: Elizabeth Buie. Setting: Washington DC.
#31: Get some input if you can.
Ideally, a colleague would use the technique independently, and then youd consolidate results. of Two Friends. From Wikimedia Commons: http://en.wikipedia.org/wiki/File:Jacopo_Pontormo_-_Portrait_of_Two_Friends_-_WGA18109.jpg
#32: Maybe you can get someone to review your tagging, grouping and naming.
Artwork: Jacopo Pontormo, Portrait of Two Friends. From Wikimedia Commons: http://en.wikipedia.org/wiki/File:Jacopo_Pontormo_-_Portrait_of_Two_Friends_-_WGA18109.jpg
#33: Failing even that, see if you can have someone do independent tagging, and add their keywords to your list before you start grouping.
Artwork: Jacopo Pontormo, Portrait of Two Friends. From Wikimedia Commons: http://en.wikipedia.org/wiki/File:Jacopo_Pontormo_-_Portrait_of_Two_Friends_-_WGA18109.jpg
#34: Now you have a draft IA that you can try out. It may not be as good as what you could have obtained with a user study, but its a start.
Id like to see someone try card sorting with tags rather than larger bits of content, and see what happens.
#36: Photo credit: Elizabeth Buie. Setting: Halifax, Nova Scotia, Canada