R help - Sep 2015 - Composite index reliability questions - cronbach()

Hi All,

I have two questions on using cronbach() from the psy() package.

My simplified situation is the following: I have a survey of 10 questions
(column names are "Q1", "Q2", etc.) that went out to 100+
people. I have
the responses to the questions, plus additional variables (demographics,
location, etc.) in a data frame (named "JDC"). I want to build
composite
indices from these 10 questions. I plan to use two steps to create the
indices. First, grouping the questions by what makes intuitive sense given
what they ask, and second, by testing the reliability of these groupings
using cronbach().

QUESTION 1

Let's say I think Q1, Q3, and Q5 will make a good index. With my limited
knowledge of R, I would think there's a way to say "run the reliability
on
these three variables in this dataframe". However, I have so far only been
able to test the reliability of *adjacent *variables. For example, I could
do:

*cronbach(jdc[,1:3])*

to test Q1, Q2, and Q3. Is there a way to test non-adjacent variables?

I realize I could do something like:

*trust <- jdc[, c("Q2", "Q7", "Q8")]*
*cronbach(trust)*

but that adds a few extra steps, and I have tons of questions and indices
which would make that very cumbersome, especially since I will go through
several iterations in testing potential indices.


QUESTION 2

Is there a way to refer to the column name when using cronbach(), instead
of just the location of the variable? For example:

This works:   *cronbach(jdc[,1:3])*
This doesn't:* cronbach(jdc[Q1, Q2, Q3])*


Thanks in advance for any insights, answers, words of encouragement, or
alternate ways I could solve this puzzle.

Nick

	[[alternative HTML version deleted]]

Hi Nick,
If you haven't just made a typo on your example in QUESTION 2, the
"This
doesn't" line should read:

cronbach(jdc[,c("Q1","Q2","Q3")])

Without the quotes, R looks for three objects named Q1, Q2 and Q3 and
probably doesn't find them.

Jim


On Sat, Sep 5, 2015 at 7:27 AM, Nick Petschek <nick.petschek at gmail.com>
wrote:
> Hi All,
>
> I have two questions on using cronbach() from the psy() package.
>
> My simplified situation is the following: I have a survey of 10 questions
> (column names are "Q1", "Q2", etc.) that went out to
100+ people. I have
> the responses to the questions, plus additional variables (demographics,
> location, etc.) in a data frame (named "JDC"). I want to build
composite
> indices from these 10 questions. I plan to use two steps to create the
> indices. First, grouping the questions by what makes intuitive sense given
> what they ask, and second, by testing the reliability of these groupings
> using cronbach().
>
> QUESTION 1
>
> Let's say I think Q1, Q3, and Q5 will make a good index. With my
limited
> knowledge of R, I would think there's a way to say "run the
reliability on
> these three variables in this dataframe". However, I have so far only
been
> able to test the reliability of *adjacent *variables. For example, I could
> do:
>
> *cronbach(jdc[,1:3])*
>
> to test Q1, Q2, and Q3. Is there a way to test non-adjacent variables?
>
> I realize I could do something like:
>
> *trust <- jdc[, c("Q2", "Q7", "Q8")]*
> *cronbach(trust)*
>
> but that adds a few extra steps, and I have tons of questions and indices
> which would make that very cumbersome, especially since I will go through
> several iterations in testing potential indices.
>
>
> QUESTION 2
>
> Is there a way to refer to the column name when using cronbach(), instead
> of just the location of the variable? For example:
>
> This works:   *cronbach(jdc[,1:3])*
> This doesn't:* cronbach(jdc[Q1, Q2, Q3])*
>
>
> Thanks in advance for any insights, answers, words of encouragement, or
> alternate ways I could solve this puzzle.
>
> Nick
>
>         [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
	[[alternative HTML version deleted]]

Thanks, Jim!

The lack of quotes was a typo, but what was not was my forgetting to
include the "c(" function... Thanks!



On Sat, Sep 5, 2015 at 7:23 AM, Jim Lemon <drjimlemon at gmail.com> wrote:
> Hi Nick,
> If you haven't just made a typo on your example in QUESTION 2, the
"This
> doesn't" line should read:
>
> cronbach(jdc[,c("Q1","Q2","Q3")])
>
> Without the quotes, R looks for three objects named Q1, Q2 and Q3 and
> probably doesn't find them.
>
> Jim
>
>
> On Sat, Sep 5, 2015 at 7:27 AM, Nick Petschek <nick.petschek at
gmail.com>
> wrote:
>
>> Hi All,
>>
>> I have two questions on using cronbach() from the psy() package.
>>
>> My simplified situation is the following: I have a survey of 10
questions
>> (column names are "Q1", "Q2", etc.) that went out
to 100+ people. I have
>> the responses to the questions, plus additional variables
(demographics,
>> location, etc.) in a data frame (named "JDC"). I want to
build composite
>> indices from these 10 questions. I plan to use two steps to create the
>> indices. First, grouping the questions by what makes intuitive sense
given
>> what they ask, and second, by testing the reliability of these
groupings
>> using cronbach().
>>
>> QUESTION 1
>>
>> Let's say I think Q1, Q3, and Q5 will make a good index. With my
limited
>> knowledge of R, I would think there's a way to say "run the
reliability on
>> these three variables in this dataframe". However, I have so far
only been
>> able to test the reliability of *adjacent *variables. For example, I
could
>> do:
>>
>> *cronbach(jdc[,1:3])*
>>
>> to test Q1, Q2, and Q3. Is there a way to test non-adjacent variables?
>>
>> I realize I could do something like:
>>
>> *trust <- jdc[, c("Q2", "Q7", "Q8")]*
>> *cronbach(trust)*
>>
>> but that adds a few extra steps, and I have tons of questions and
indices
>> which would make that very cumbersome, especially since I will go
through
>> several iterations in testing potential indices.
>>
>>
>> QUESTION 2
>>
>> Is there a way to refer to the column name when using cronbach(),
instead
>> of just the location of the variable? For example:
>>
>> This works:   *cronbach(jdc[,1:3])*
>> This doesn't:* cronbach(jdc[Q1, Q2, Q3])*
>>
>>
>> Thanks in advance for any insights, answers, words of encouragement, or
>> alternate ways I could solve this puzzle.
>>
>> Nick
>>
>>         [[alternative HTML version deleted]]
>>
>> ______________________________________________
>> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide
>> http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
>>
>
>
	[[alternative HTML version deleted]]