| TESTING THE SPANISH VERSION 1999 |
| With Maria Luisa Chavez, Universidad de Nuevo Leon, Monterrey, Mexico |
| Data with the translated ASF-E [24-item version tested with sample 6 (see Development of the ASF-E) were collected from 258 Mexican families in extreme and moderate poverty. Participants were 133 adolescents and 125 mothers, fathers and other adults in charge of the family. A second version produced in Mexico with similar questions based on the same framework was also tested with 407 comparable families (222 adolescents and 185 mothers, fathers or other adults in charge of the family). Factor analyses were done on both instruments. Fourteen solid items with acceptable reliability resulted from the analysis of the translation and 17 from the second instrument. This tested version was a combination of all 31 items, one additional item was added. |
| Subjects |
| Adolescents were 50% girls and 50% boys. All were students in the secondary school, in first to third grade. Most were students only, but 9% also carried a job. Of the adolescents, 64% were 12 to 14 years old and 36% were between 15 and 20. |
| Other family members (mothers, fathers, siblings, etc.) were predominantly women (87%). Six percent were less than 20 years old, 28% between 21 and 35 years and 66% were between 36 and 64 years. Ten percent were single, 76% married, 7% living together and 8% widowed or other. Their educational level was low with 8% having no schooling, 41% had primary school only, 25% secondary school, 8% preparatory school and 18% some other type of education. The majority (68%) tended the household, 25% carried a job, 5% were students and 2% worked while studying. |
| Item testing |
| Six items were excluded initially due to skewed distributions. They had p-values greater than .80. |
| Factor Analyses |
| Initial Exploratory Factor Analyses (PC analysis with Varimax orthogonal rotation) |
| Analyses were done for all subjects, as well as for the teens and the adults separately. The remaining 26 items were factor analyzed. There were 8 factors for the adults, 10 for the teens and 7 for the whole group. Items were then examined within each factor across the three analyses. Items that grouped together in the same factor in at least two of the analyses were retained, to assure their relevance in the two distinct groups of subjects. All items were retained. |
| Further analyses were conducted with the total sample. Factors were limited to 7,6,5, and 4 factors. Additional items were excluded if their factor loadings were smaller than .40. |
| Final Factor Analysis |
| The best version was a 4-factor solution with a total of 20 items. Eigenvalues ranged from 4.51 to 1.17. The four factors explained 41.7% of the variance. |
| The factor solution was theoretically sound: |
| SM = | System Maintenance |
| C = | Coherence |
| SC = | System Change |
| I = | Individuation |
| Item Description | Factors |
| F 1 | F 2 | F 3 | F 4 | |
| SM | SC | C | I | |
| Mutual help to get tasks done | 0.66 | |||
| Tasks for males and females | 0.57 | |||
| Recreation together | 0.57 | |||
| Happiness/sadness in the family | 0.53 | |||
| Agreement about division of labor | 0.50 | |||
| Helping out in difficult times | 0.48 | |||
| Effect of problems | 0.62 | |||
| Introduction of new ideas | 0.61 | |||
| Acceptance of opinions | 0.57 | |||
| Talk about things members learn | 0.50 | |||
| Knowing what happens at work, school | 0.46 | |||
| Solving problems | 0.43 | |||
| Holding together to straighten out others | 0.64 | |||
| Comfort with neighbors | 0.59 | |||
| Openness about personal views | 0.52 | |||
| Feeling about the family | 0.48 | |||
| Mutual understanding | 0.43 | |||
| Work in the community | 0.69 |
|||
| Participation in community activities | 0.66 | |||
| Engage in favorite activities. | 0.64 | |||
| InternalConsistency (Cronbach’s alpha) | 0.64 | 0.67 | 0.55 | 0.55 |
| Internal Consistency - Reliability (Cronbach’s alpha) |
| Items loading on the same factor (process dimensions) are listed above. |
| Cronbach's alphas for the targets are: |
| All Subjects | Adolescents Only | Family Members Only | |
| Growth | 0.70 | 0.69 | 0.68 |
| Control | 0.77 | 0.75 | 0.76 |
| Stability | 0.71 | 0.71 | 0.66 |
| Spirituality | 0.60 | 0.60 | 0.56 |
| Total score | 0.81 | 0.80 | 0.79 |
| Most results were in an acceptable range. Family members' coefficients were lower. |
| This seems to be related to their very low education level and speaks to the necessity to assist with the completion of the instrument. Interpretation was most needed with individuation items - items that affected spirituality and stability scores. |
| Conclusion |
| The 20-item instrument has construct validity and reliability. |
| It reflects cultural differences in family perception. For example, the family is perceived as a group of people committed to each other who go out, seek information and bring in resources for the benefit of the family. Coherence is achieved by being active. Sharing information is done mainly for the sake of the family and to solve problems jointly rather than to assert one's individuality. |
| The instrument can be used with families from the low socioeconomic groups, a population that challenges any type of standardized measurement. |
| Caution needs to be exercised with people who have no or very little formal education. |
| For people who need to have the items read to them, simple standard interpretations of the individuation items would be useful. |
| The tool will need retesting for families from other socioeconomic groups. |
| THE INSTRUMENT IS AVAILABLE FROM THE AUTHORS |
321
Individuals in poverty (Region San Bernabe)
144
Low middle class families (Region Morelos)
Demographics:
Adults
174 (Mothers 145, Fathers 11, Others 18)
Teens
291
Gender:
295 Female
170 Male
Age:
Adolescents: 11-15 years
Adults:
35-64 years
Marital
Status of Adults:
18
Single
132 Married
10 Living
together
14 Other
Education
of Adults:
13 None
72 Primary
School
44 Secondary
School
14 Preparatory
School
31 Other
Occupation
of Adults:
119 Housekeepers
45 Employed
9
Work and study
Members in the
Family
2
– 5
57 %
6
and more 43 %
Type of Family
Nuclear
family
75%
Parents,
children, relatives
22%
Parents,
children, relatives,
non-relatives
3%
The
factor structure upon exploratory analysis was almost identical to the above.
Reliabilities
(Cronbach
Alpha)
.70 for Growth
.70 for Stability
.60 for Spirituality
.77 for Control
For the adolescents alone, and adults alone, reliabilities are similar.
No significant differences between levels of social status.
Income: Less than $10,000
44%
$10,000 – 30,000
38 %
more than $30,000 3 %
Unknown
15
%
Demographics:
Gender:
735 Girls
649 Boys
Numbers of
Family Members:
2 – 5
77 %
6
and more 23 %
Results:
Tested with reliability analysis. Even though no adults were included, Cronbach alpha coefficients were satisfactory:
.71
for Growth
.64
for Spirituality
.73
for Stability
.77
for Control
The results were very comparable for lower and
higher incomes.
Conclusion:
Instrument holds firm for the various populations tested. Response
distribution was problematic for one item only (p > .90).