A Pilot Test of the Midlife Womens Symptom Index MSI

1
A Pilot Test of the Midlife Womens Symptom
Index (MSI)

• Eun-Ok Im, PhD, MPH, RN, CNS
• Wonshik Chee, PhD
• The Univ. of Texas at Austin
• The University of Texas at San Antonio

2
Purpose of the Study

• To determine reliability and validity of the
  Midlife Womens Symptom Index (MSI)
• To compare the pen-and-pencil format and Internet
  format of the MSI.

3
Theoretical Framework

• Feminist Approach

4
Study Design

• Cross-sectional comparative design with three
  sections
• Group 1 a pen-and-pencil questionnaire survey
• Group 2 an Internet survey
• Group 3 an expert review

5
Group 1

• The 77 women recruited through settings such as
  shopping malls, beauty shops, and churches
  included 6 Hispanic, 50 N-H White, 4 N-H
  African-American, and 17 N-H Asian women.

6
Group 2

• The 81 women recruited through Internet
  communities/groups (ICs) and flyers included 17
  Hispanic, 44 N-H White, 9 N-H African-American,
  and 11 N-H Asian women.

7
Group 3

• Twenty experts in womens health who were faculty
  members of schools of nursing at 5 universities
  were recruited using a convenience sampling
  method (Group 3).
• This group of womens health experts included 13
  N-H whites, 2 N-H African-Americans, 4 N-H
  Asians, and 1 N-H Hispanic.

8
Instruments

• Both pen-and-pencil questionnaire and Internet
  questionnaire included self-administered
  questions on 19 sociodemographic characteristics
  including ethnic identity and menopausal status,
  the Menopausal Symptom Index (MSI), and the
  Menopausal Symptom Checklist (MSC).
• For the expert review, only one Likert-scale
  question on face validity was used.

9
Data Collection Procedures

• Group 1 was asked to self-administer the
  pen-and-pencil questionnaire and return the
  questionnaire by mail or in person.
• Group 2 was asked to visit the project website
  and fill out the questionnaire through the
  Internet.
• Group 3 was asked to review the MSI, answer the
  question on face validity, and provide comments
  in an open space.

10
Data Analysis Process

• Sociodemographic Characteristics and Symptom
  Experience of Groups 1 and 2 Descriptive
  Statistics and Item Analysis.
• Convergent validity Correlations between total
  scores of total, physical, psychological, and
  psychosomatic symptoms of the MSI to the
  corresponding scores of the MSC.
• Face validity Descriptive statistics.
• Item omission, response rate, and completeness
  Descriptive statistics.

11
Findings Ethnic Differences

• The mean total number of symptoms experienced by
  Group 2 (mean25.6, SD10.29) was larger than
  that by Group 1 (mean22.5, SD14.69), but the
  difference was not statistically significant.

12
Findings Ethnic Differences of Group 1

• There was an ethnic difference in the total
  number of total symptoms (F2.36, p0.09).
  Chi-square tests of the frequency of each symptom
  by ethnicity also indicated ethnic differences in
  the frequency of some menopausal symptoms
  including sore throat, poor appetite, upset
  stomach, stiff and sore joints, feeling hot or
  cold, twitching of face, head or shoulder, tense
  or jumpy during menstruation, hot flashes, sleep
  disturbances, breast pain, vaginal dryness,
  forgetfulness, and bleeding between periods
  (plt.05).

13
Findings Ethnic Differences in Group 2

• In Group 2, there was no ethnic difference in the
  total number of total symptoms. Yet, chi-square
  tests of the frequency of each symptom by
  ethnicity indicated ethnic differences in the
  frequency of some menopausal symptoms including
  vision changes, watery eyes, feeling bloated,
  pain in stomach, dizziness, numbness or tingling
  sensation, frequent urination, flooding, and
  itching on genital area (plt0.05).

14
Findings Reliability of MSI

• The findings supported the internal consistency
  reliability of the pen-and-pencil and the
  Internet format of MSI. K-R20 for the
  pen-and-pencil format was 0.94 among all
  participants, 0.94 among Whites, 0.88 among
  Asians, and 0.97 among Hispanics.
• K-R20 for the pen-and-pencil format among African
  Americans could not be calculated because of the
  small sample size of this subgroup (n4).
• K-R20 for the Internet format was 0.86 among all
  the participants, 0.89 among Whites, 0.78 among
  Hispanics, and 0.86 among Asians. K-R20 for the
  Internet format among African Americans could not
  be also calculated because of the small sample
  size (n9).

15
Findings Convergent Validity

• For the pen-and-pencil format, the correlation
  coefficient between the total number of symptoms
  of MSI and MSC was 0.80.
• For the Internet format, the correlation
  coefficient between the total number of symptoms
  of MSI and MSC was 0.94.

16
Findings Convergent Validity of Pen-and-Pencil
Format

• Correlation coefficient between the total number
  of physical symptoms was 0.80 that between the
  total number of psychological symptoms was 0.93
  and that between the total number of
  psychosomatic symptoms was 0.75.

17
Findings Convergent Validity of Internet Format

• Correlation coefficient between the total number
  of physical symptoms was 083 that between the
  total number of psychological symptoms was 0.61,
  and that between the total number of
  psychosomatic symptoms was 0.63.

18
Findings Item Analysis

• Based on the item analysis and suggestions from
  the expert review, 15 items that had the
  inter-item correlations greater than 0.70 or less
  than 0.30 and that were found to be inappropriate
  were eliminated. After removing the items, K-R20
  was not significantly changed (0.94 for the
  pen-and-pencil format and 0.87 for the Internet
  format), but the item-total correlations and the
  inter-item correlations were improved.

19
Discussion 1

• The findings suggest that there are certain
  ethnic differences in symptom experience during
  the menopausal transition.

20
Discussion 2

• The findings suggest that the Internet survey can
  produce comparable data to the pen-and-pencil
  survey.

21
Discussion 3

• The findings suggest that Internet survey is
  better in recruiting multi-ethnic groups of
  research participants.

22
Discussion 4

• The findings support that there is no significant
  difference in psychometric properties of the
  pen-and-pencil format and the Internet format of
  the MSI.

23
Implications

• Further studies on ethnic differences in symptom
  experience during the menopausal transition are
  needed.
• Further comparisons of psychometric properties of
  the Internet format and the pen-and-pencil format
  of a questionnaire are needed.
• Further efforts need to be made to explore if
  Internet research can produce comparable data to
  traditional research methods.