Considering the Integration of Qualitative and Quantitative

1
Considering the Integration of Qualitative and
Quantitative

• A Test of AAHSL Stats and LibQUAL Data
• Doug Joubert, Lyn Dennison and Tamera Lee
• Medical College of Georgia

2
Local Questions

• Do any patterns exist between AAHSL Annual Stats
  and LibQUAL data?
• Required combining the data from both data sets
  into a common SPSS file

3
Local Questions

• AAHSL Data Transformation
• Examined the Expenditures Summary data from AAHSL
  Annual Statistics
• Transformed and recoded the AAHSL data to
  accommodate for missing scores
• For example, with data import, SPSS needed to
  understand that M was a system missing value

4
Local Questions

• AAHSL Data Transformation
• Coded AAHSL variables
• Personnel expenditures?perexp
• Total Collection Expenditures ?tocex
• Total Recurring Expenditures ?toreex
• Capital Budget ?capbud
• Total Annual Expenditures ?toanex

5
Local Questions

• AAHSL Data Transformation
• Combined the information from both data sets into
  a single SPSS data file
• Grouped data by a common variable instID
• Merged the two files via SPSS

6
Local Questions

• SPSS Data Transformation
• Identified stats for the Affect of Service
  Dimension
• LibQUAL already created variables for person
  level subscales
• Specifically, the minimum, desired, and perceived
  means for each of the 2002 Dimensions
• Computed the means of Service Affect Dimension
  for each participating institution

7
Local Questions

• SPSS Data Transformation
• To compute the means of Service of Affect
  Dimension for institutions we used the following
  variables from LibQUAL
• aavgmin1
• aavgdes1
• aavgper1

8
Local Questions

• SPSS Data Transformation
• Computing the means of Service of Affect
  Dimension for each institution allowed us to
  compute a mean gap for each institution
• This was accomplished in much the same way as
  computing the LibQUAL gap
• Average perceived average minimum average gap
  (by institution)

9
Local Questions

• SPSS Data Transformation
• Having the average gap (by institution) allowed
  us to look at the relationship between it and
  the total annual expenditures (AAHSL)

10
(No Transcript)
11
Local Questions

• SPSS Data Transformation
• Having the average gap (by institution) also
  allowed us to transform the gap score into a
  T-score (Norm Table)
• As discussed Cook et al., T-scores allow one to
  examine individuals scores in relation to scores
  of peer insitutions1

12
Norms based on Affect of Service Gap
13
Local Questions

• Questions for further exploration
• The scatter plot visually reveals no relationship
  between gap score and Total Annual Expenditures

14
Local Questions

• Questions for further exploration
• Q 1 What valid statistical method may be used to
  measure correlation with the gap score?
• For example Spearman rank-order, Pearson
  correlation, or Linear Regression
• Q 2 How do we develop percentile ranks based on
  T-scores (norms) in any number of questions and
  dimensions?

15
References

1. Cook, C., Heath, H., and Thompson, B. Score Norms
   for Improving Library Service Quality A LibQUAL
   Study. portal Libraries and the Academy, vol. 2,
   no. 1, pp. 13-26. (2002)