Help: Sampling for qualitative data mining

Thread View: bit.listserv.qualrs-l

2 messages

[T] Back to Threads [V] Tree View [G] Group Articles

2 total messages Started by Anthony Mon, 18 Feb 2002 23:39

#779

Author: Anthony
Date: Mon, 18 Feb 2002 23:39

23 lines
2019 bytes

Newbie question...forgive me if I use incorrect terminology

I work for a software company. I'm analyzing notes from calls that
customers' made to our technical support department. My goal is to get
a better understanding of what's causing the customers to call.

Basically, my approach is to read the notes, which are an summary of
the call conversation, code each call by a the problems that caused
the customer to call, and look for patterns in the data.

My question is this: Can someone recommend a sampling strategy so that
I can read and code a subset of the calls, yet still feel confident
that the sample I coded is representative of the rest of the calls?
For example, if I find that half of the calls I analyzed were caused
because customers couldn't understand a particular error message, I
want to be able to feel confident that the roughly half of the rest of
the calls were caused by that problem.

I know that there are a lot of issues that could be discussed about
this, but my main question is about how get a good sample. Any help
will be appreciated.

Anthony

Re: Sampling for qualitative data mining

#784

Author: Joao
Date: Tue, 09 Apr 2002 15:47

36 lines
2456 bytes

Anthony,

How many messages do you have?

One alternative is to use a qualitative software package and code them
automatically...

Joao



"Anthony" <azablocki@yahoo.com> wrote in message
news:a6e86857.0202181539.573de553@posting.google.com...
> Newbie question...forgive me if I use incorrect terminology
>
> I work for a software company. I'm analyzing notes from calls that
> customers' made to our technical support department. My goal is to get
> a better understanding of what's causing the customers to call.
>
> Basically, my approach is to read the notes, which are an summary of
> the call conversation, code each call by a the problems that caused
> the customer to call, and look for patterns in the data.
>
> My question is this: Can someone recommend a sampling strategy so that
> I can read and code a subset of the calls, yet still feel confident
> that the sample I coded is representative of the rest of the calls?
> For example, if I find that half of the calls I analyzed were caused
> because customers couldn't understand a particular error message, I
> want to be able to feel confident that the roughly half of the rest of
> the calls were caused by that problem.
>
> I know that there are a lot of issues that could be discussed about
> this, but my main question is about how get a good sample. Any help
> will be appreciated.
>
> Anthony

Thread Navigation

This is a paginated view of messages in the thread with full content displayed inline.

Messages are displayed in chronological order, with the original post highlighted in green.

Use pagination controls to navigate through all messages in large threads.

Back to All Threads