The Assessment of Strategies

       In Families - Effectiveness

(ASF-E)

______________________________________________________
______________________________________________________
      Development Of The ASF-E
Method:
*Development of an item pool by a group of professionals from various social sciences and family service agencies.
*Revision and validation of items (content validity): Independent rating of items by five family experts; exclusion of items not unanimously rated.
*Item administration: Tool completed by adult family members (one per family) from a community sample of 622.
*Exploratory factor analysis (PC analysis with Varimax orthogonal rotation)
*Selection of best items for further testing (item reduction) (construct validity)
*Retesting items with community and special samples
*Reliability testing (internal consistency)
*Discriminant validity testing (clinical versus community subjects)
Samples:
1) Community sample N=622.
                           Age Range: 18-90 years
                          Gender:   Females  68%
                                          Males     32%
       
                              Race:       White     56%
                                               Black     37%
                                               Others     7%
         Family Type:     Adult and Elderly Families                   40%
                                 Families with Adolescents andchildren  60%                                            
                                    Clinical Families                                   37%
       (mental  illness, domestic violence, substance abuse)    
                                  Community Families                               63%
                                   
2) Community sample N=52
                 Age Range:    30-57 years
                             Gender:     Females   50%
                                                Males      50%
                             Race:        All White
         Family Type:     Adults and Elderly Families                    35%
                                  Families with Adolescents and Children 65%
                                      
                                    No Clinical Families.
3) Clinical sample N=117
    Substance abuse residential treatment.
    Residents (90% male) with one family member each.
                              Mean Age:     40.5years
                              Gender:     Females      49%
                                               Males         51%
                              Ethnicity:   Black          80%
                                               White        14%
                                               Other           6%
         Family Type:     Adult and Elderly Families      60%
                                 Families with Children              40%
                               Income:    < $10,000              40% 
                                                $10,000-25,000    34% 
                                                > $25,000              19%    
                                                Missing                    7%
               Education:  Less than High School            28%
                                   High School and More       72%
4) Community sample N = 237
    Relatively young, well educated
                             Ethnicity:      White         67%
                                                 Black         23%
                                                 Other         10%
5) Families with chronic pain N = 30
 Pain lasting from 1 to 43 years (average 9 years).
                         Age Range        31 - 82 years
                               Ethnicity:    White       83%
                                                 Black.     17%
                        Education:       Well-educated
                               Income:         Moderate
6) Community sample N = 445
                            Age range         18 - 81years
                            Mean Age          36 years
                            Gender:       Females      65%
                                               Male           35%
                           Ethnicity:      White                    55%
                                              Black                    14%
                                              Hispanic               12%
                                              Asian                    11%
                                              Native American     5%
                                              Other                      3%
         Family Type:        Adult Families                     55%
                                    With Children < 18 years   44%
                                    Missing                                1%
                                   Two Parents                        53%
                                   Single Parent                       18%
                                   One Generation                   28%
                                   Missing                                  1%
                        Income:      < $10,000                11%
                                            $10,000 - 25,000    23%
                                           > $25,000                66%
               Education:      Less than High School           4%
                                     High School                          42%
                                     College                                 54%
Results:
Factor Analysis yielded 4 factors initially:

 

  Stability (System maintenance and coherence)     6 items
  Growth  (System change and individuation)         6 items 
  External Support                                                 3 items.
Internal Consistency -Reliability
            Cronbach's alpha .  Subscales    .60 to .80
                                 Total Score:                  .84
Details, see Friedemann, 1991.
In all subsequent samples, external support did not factor
as an independent dimension, especially with minority
subjects, but was folded into the others. This was
consistent with the framework. Therefore, these items were
assigned to System Change or System Maintenance.
Factor structure was indicative of all four dimensions. In
each testing, a few items were added, some taken off. In all
 test runs, system maintenance and coherence factored
together. Items were added to the Growth dimensions for
better reliability; reliability for Growth ranged from .50 to .62
 in various test runs.
Test-Retest Reliability
with 43 substance abusing families (sample 3) was
satisfactory:
                      Stability at    1 month       r = .94
                                          2 months     r = .85
                      Growth   at   1 month       r = .80
                                          2 months      r = .56
  (Growth was expected to improve in these families)

 

                    Total   at      1 month        r= .94
                                         2 months      r = .84
Concurrent Validity with the same sample:
ASF was tested against FES (Moos). Values were
unsatisfactory. The internal consistency of the FES
with this sample was very low (.008 to .34). The FES could
not be used with minority substance abusing families.
Testing for Type III Error:
Congruence between items and the actual phenomenon -
            1) Convergence of intended and subject-ascribed
                  meaning of items
            2) Convergence of score and qualitative
                 description of family dynamics
Results very satisfactory.
Details, see Friedemann and Smith, 1997.