Factor Analysis

[Ismail Noor]

Dear All at the List,

 

My client wanted to determine the obstacles facing female faculty in three different universities, in three different countries, that hindering their achievement motivation in their scholarly research and academic development. For that, two different surveys were used to collect needed data; one for the faculty at those universities and one for the administrators. A total of 63 faculty members and 17 administrators were surveyed. Each survey was consisted of several scales to measure different factors. The faculty survey has 240 variables, including the demographic variables, and the administrators’ survey has 157 variables, including demographic variables.

In addition to the descriptive analysis, I propose to subject the total number of variables (240 for faculty, and 157 for administrators’’) to a factor analysis, as a data reduction technique, to isolate measures and factors that best evaluate the integrated model of faculty motivation (adopted from Pintrich, 1994) the researcher used. For that, a “varimax factor rotation” analysis will be employed. A stepwise regression analysis will then be conducted followed by a standard path analysis to estimate the direct and indirect effect of each measure and factor in the model that best portray the process and provide key insights into the motivation productivity of female faculty.

I am asking for a second opinion on my analysis plan. How does this analysis plan sound to you? I appreciate any thoughts on this. I forgot to state that I am using SPSS for my analysis.

 

Ismail K. Noor, Ph.D.,
[hidden email]

Research Consultanting

1335 Culver Avenue

Dearborn, Michigan 48124

[Art Kendall]

With such tiny numbers of respondents, factor analysis is inappropriate.

Did Pintarch 1994 or some other researcher have pre-existing scales that they created, validated, etc?
If so, you should just use the scale scores as the data reduction.

Art Kendall
Social Research Consultants




On 6/23/2010 8:31 AM, Ismail Noor wrote:

Dear All at the List,

�

My client wanted to determine the obstacles facing female faculty in three different universities, in three different countries, that hindering their achievement motivation in their scholarly research and academic development. For that, two different surveys were used to collect needed data; one for the faculty at those universities and one for the administrators. A total of 63 faculty members and 17 administrators were surveyed. Each survey was consisted of several scales to measure different factors. The faculty survey has 240 variables, including the demographic variables, and the administrators’ survey has 157 variables, including demographic variables.

In addition to the descriptive analysis, I propose to subject the total number of variables (240 for faculty, and 157 for administrators’’) to a factor analysis, as a data reduction technique, to isolate measures and factors that best evaluate the integrated model of faculty motivation (adopted from Pintrich, 1994) the researcher used. For that, a “varimax factor rotation” analysis will be employed. A stepwise regression analysis will then be conducted followed by a standard path analysis to estimate the direct and indirect effect of each measure and factor in the model that best portray the process and provide key insights into the motivation productivity of female faculty.

I am asking for a second opinion on my analysis plan. How does this analysis plan sound to you? I appreciate any thoughts on this. I forgot to state that I am using SPSS for my analysis.

�

Ismail K. Noor, Ph.D.,
[hidden email]

Research Consultanting

1335 Culver Avenue

Dearborn, Michigan 48124

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

[statisticsdoc]

Ismail,
In order to carry out factor analysis, you should have multiple subjects per variable rather than multiple variables per subject. If you want to reduce the number of variables in the study, apply scoring systems that have been previouly validated by investigators based upon larger samples, as Art has suggested.
HTH,
Stephen Brand, Ph.D.

www.StatisticsDoc.com

---

From: Ismail Noor <[hidden email]>

Sender: "SPSSX(r) Discussion" <[hidden email]>

Date: Wed, 23 Jun 2010 05:31:03 -0700

To: <[hidden email]>

ReplyTo: Ismail Noor <[hidden email]>

Subject: Factor Analysis

Dear All at the List,

 

My client wanted to determine the obstacles facing female faculty in three different universities, in three different countries, that hindering their achievement motivation in their scholarly research and academic development. For that, two different surveys were used to collect needed data; one for the faculty at those universities and one for the administrators. A total of 63 faculty members and 17 administrators were surveyed. Each survey was consisted of several scales to measure different factors. The faculty survey has 240 variables, including the demographic variables, and the administrators’ survey has 157 variables, including demographic variables.

In addition to the descriptive analysis, I propose to subject the total number of variables (240 for faculty, and 157 for administrators’’) to a factor analysis, as a data reduction technique, to isolate measures and factors that best evaluate the integrated model of faculty motivation (adopted from Pintrich, 1994) the researcher used. For that, a “varimax factor rotation” analysis will be employed. A stepwise regression analysis will then be conducted followed by a standard path analysis to estimate the direct and indirect effect of each measure and factor in the model that best portray the process and provide key insights into the motivation productivity of female faculty.

I am asking for a second opinion on my analysis plan. How does this analysis plan sound to you? I appreciate any thoughts on this. I forgot to state that I am using SPSS for my analysis.

 

Ismail K. Noor, Ph.D.,
[hidden email]

Research Consultanting

1335 Culver Avenue

Dearborn, Michigan 48124

[Art Kendall]

Each scale item is reflected if necessary and then summed or averaged into a scale score.
After you check on the keying via RELIABILITY, you just let the scale scores represent all the interesting variance in the items that were used to construct them.

That� greatly reduces the number of variables that would be used in any analysis.

Art Kendall
Social Research Consultants


On 6/26/2010 1:50 AM, Ismail Noor wrote:

Thank you for your� help. Actually, the two surveys used in the study were adopted from� Robert T. Blackbarn & Janet Larence (forgot the year) book; Faculty at work: motivation, expectation, satisfaction, with very minor adjustment of very few questions. So, the scales used are pre-existing scales and I think they were validated.

�

But could you please elaborate on the comments you made and tell me how would use the scale scores as the data reduction? Do you need more information on the data the researcher collected, or copy of the survey?

�

Thank you and best regards..............
�

Ismail K. Noor, Ph.D.,
[hidden email]

Research Consultant

1335 Culver Avenue

Dearborn, Michigan 48124

---

From: Art Kendall [hidden email]
To: [hidden email]
Sent: Wed, June 23, 2010 11:49:41 PM
Subject: Re: Factor Analysis

With such tiny numbers of respondents, factor analysis is inappropriate.

Did Pintarch 1994 or some other researcher have pre-existing scales that they created, validated, etc?
If so, you should just use the scale scores as the data reduction.

Art Kendall
Social Research Consultants




On 6/23/2010 8:31 AM, Ismail Noor wrote:

Dear All at the List,

Â

My client wanted to determine the obstacles facing female faculty in three different universities, in three different countries, that hindering their achievement motivation in their scholarly research and academic development. For that, two different surveys were used to collect needed data; one for the faculty at those universities and one for the administrators. A total of 63 faculty members and 17 administrators were surveyed. Each survey was consisted of several scales to measure different factors. The faculty survey has 240 variables, including the demographic variables, and the administrators’ survey has 157 variables, including demographic variables.

In addition to the descriptive analysis, I propose to subject the total number of variables (240 for faculty, and 157 for administrators’’) to a factor analysis, as a data reduction technique, to isolate measures and factors that best evaluate the integrated model of faculty motivation (adopted from Pintrich, 1994) the researcher used. For that, a “varimax factor rotation� analysis will be employed. A stepwise regression analysis will then be conducted followed by a standard path analysis to estimate the direct and indirect effect of each measure and factor in the model that best portray the process and provide key insights into the motivation productivity of female faculty.

I am asking for a second opinion on my analysis plan. How does this analysis plan sound to you? I appreciate any thoughts on this. I forgot to state that I am using SPSS for my analysis.

Â

Ismail K. Noor, Ph.D.,
[hidden email]

Research Consultanting

1335 Culver Avenue

Dearborn, Michigan 48124

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

[Mark A Davenport MADAVENP]

Ismail,

Are you referring Paul Pintrich and the MSLQ?  If so, there has been a good deal of validity and reliability work done with that instrument as I learned when I did my dissertation--and that was back in 1999:

Modeling motivation and learning strategy use in the classroom: An assessment of the factorial, structural, and predictive validity of the Motivated Strategies for Learning Questionnaire. Ph.D. dissertation, Auburn University, United States -- Alabama. Retrieved June 30, 2010, from Dissertations & Theses: Full Text.(Publication No. AAT 3081568).

If you used the motivation scales of that instrument intact and without changes to the items, you should find plenty of information to help you make judgments about creating factor scores.  Predictably, I found the motivation scales to be a mixed bag for the sample I was using--the constructs with four or fewer items loading on a factor had marginal reliability.  Granted, my sample was fairly small and homogeneous sample (226 cases).  There is plenty of available evidence with numerous other samples.

Mark

***************************************************************************************************************************************************************
Mark A. Davenport Ph.D.
Senior Research Analyst
Office of Institutional Research
The University of North Carolina at Greensboro
336.256.0395
[hidden email]

'An approximate answer to the right question is worth a good deal more than an exact answer to an approximate question.' --a paraphrase of J. W. Tukey (1962)

-----"SPSSX(r) Discussion" [hidden email] wrote: -----

To: [hidden email]
From: Art Kendall [hidden email]
Sent by: "SPSSX(r) Discussion" [hidden email]
Date: 06/30/2010 12:35PM
Subject: Re: Factor Analysis

Each scale item is reflected if necessary and then summed or averaged into a scale score.
After you check on the keying via RELIABILITY, you just let the scale scores represent all the interesting variance in the items that were used to construct them.

That? greatly reduces the number of variables that would be used in any analysis.

Art Kendall
Social Research Consultants


On 6/26/2010 1:50 AM, Ismail Noor wrote:

Thank you for your? help. Actually, the two surveys used in the study were adopted from? Robert T. Blackbarn & Janet Larence (forgot the year) book; Faculty at work: motivation, expectation, satisfaction, with very minor adjustment of very few questions. So, the scales used are pre-existing scales and I think they were validated.

?

But could you please elaborate on the comments you made and tell me how would use the scale scores as the data reduction? Do you need more information on the data the researcher collected, or copy of the survey?

?

Thank you and best regards..............
?

Ismail K. Noor, Ph.D.,
[hidden email]

Research Consultant

1335 Culver Avenue

Dearborn, Michigan 48124

---

From: Art Kendall [hidden email]
To: [hidden email]
Sent: Wed, June 23, 2010 11:49:41 PM
Subject: Re: Factor Analysis

With such tiny numbers of respondents, factor analysis is inappropriate.

Did Pintarch 1994 or some other researcher have pre-existing scales that they created, validated, etc?
If so, you should just use the scale scores as the data reduction.

Art Kendall
Social Research Consultants




On 6/23/2010 8:31 AM, Ismail Noor wrote:

Dear All at the List,

Â

My client wanted to determine the obstacles facing female faculty in three different universities, in three different countries, that hindering their achievement motivation in their scholarly research and academic development. For that, two different surveys were used to collect needed data; one for the faculty at those universities and one for the administrators. A total of 63 faculty members and 17 administrators were surveyed. Each survey was consisted of several scales to measure different factors. The faculty survey has 240 variables, including the demographic variables, and the administrators’ survey has 157 variables, including demographic variables.

In addition to the descriptive analysis, I propose to subject the total number of variables (240 for faculty, and 157 for administrators’’) to a factor analysis, as a data reduction technique, to isolate measures and factors that best evaluate the integrated model of faculty motivation (adopted from Pintrich, 1994) the researcher used. For that, a “varimax factor rotation� analysis will be employed. A stepwise regression analysis will then be conducted followed by a standard path analysis to estimate the direct and indirect effect of each measure and factor in the model that best portray the process and provide key insights into the motivation productivity of female faculty.

I am asking for a second opinion on my analysis plan. How does this analysis plan sound to you? I appreciate any thoughts on this. I forgot to state that I am using SPSS for my analysis.

Â

Ismail K. Noor, Ph.D.,
[hidden email]

Research Consultanting

1335 Culver Avenue

Dearborn, Michigan 48124

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD

====================To manage your subscription to SPSSX-L, send a message to [hidden email] (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD