Family Environment Scale: Name and Type of Measure
Family Environment Scale: Name and Type of Measure
Family Environment Scale: Name and Type of Measure
cohesion, conflict, and expressiveness, which (2009) reported 2-month test-retest reliabilities
may be examined individually or as a single ranging from 0.68 to 0.86 (M = 0.78). In addition,
index. All three of these scales are associated stability over a period of 1 year ranged from 0.53
with important outcomes. Low family conflict to 0.84 (M = 0.63), and many of the subscales
and high expressiveness and cohesion are associ- remained stable even after 10 years with a range
ated with lower health risk behavior, less from 0.38 to 0.77 (M = 0.48). The most stable
externalizing behavior in children, and greater subscale overall is moral-religious orientation,
physical fitness and psychological health (e.g., while the least stable is conflict.
Kronenberger and Thompson 1990; Nock et al. One potential psychometric limitation of the
2016; Sandhu 2015). FES is that there is little research supporting factor
validity. The instrument contains 90 items which
are presumed to be indicators of 10 different
Psychometrics scales. It is important to demonstrate that each
scale is comprised of a coherent set of items
The FES was normed using data from non- that are all good indicators of a single construct
distressed and distressed families. Families were and that clear distinctions can be made between
recruited from across the United States, with the scales. Although several studies have investi-
intentional recruiting of special-interest families, gated the higher-order factor structure of the ten
including large families, single-parent families, scales, much less is known about the factor struc-
and Latino and African-American families. Inde- ture of the items, which is essential for demon-
pendent investigators have also collected data strating factor validity. One study investigating
representing individuals from normal and dis- item factor structure produced mixed results
tressed families, and the aggregate individual (Sanford et al. 1999). On the one hand, this
data have been found to be comparable to the study identified four scales with items that failed
previously gathered family means for normal to fit the expected factor structure: expressiveness,
and distressed families (Moos and Moos 2009). independence, achievement orientation, and con-
In developing the FES, efforts were made to trol. On the other hand, after dropping these prob-
maintain high content validity for each scale, even lematic scales, a confirmatory factor analysis with
if this required inclusion of heterogeneous items the remaining six scales produced a fit that was
that lowered estimates of reliability based on judged to be adequate. Similar results were found
internal consistency. Thus, items were selected for Greek translations of the FES (Charalampous
to maximize the breadth of each construct (Moos et al. 2013). Thus, it appears that some, but not all,
and Moos 2009). Notably, some researchers have scales on the FES have adequate factor validity.
still raised questions of content validity for some Another question regarding factor structure of
scales, with one study finding disagreement the FES has to do with the extent to which scales
between experts attempting to sort items into in this instrument can be grouped together to form
their assigned scales (Roosa and Beals 1990). higher-order factors. This is important from a
Regardless, scales on the FES do appear to repre- theoretical standpoint because the scales are
sent broad constructs with estimates of internal expected to capture three broad aspects of family
consistency ranging from moderate to low. For environment: relationship, personal growth, and
example, Moos and Moos (2009) reported inter- system maintenance factors. Results from studies
nal consistency coefficients ranging from 0.61 using exploratory factor analysis typically pro-
to 0.78 (M = 0.71), and other researchers have duce results where most, but not all, scales
sometimes found slightly lower estimates (e.g., produce loadings that are consistent with the
Boyd et al. 1997). Given the breadth of items on proposed three higher-order factors. Although
each scale, a better estimate of reliability can be there has been interest in identifying higher-
gained by test-retest correlations, and these have order factors, the correlations between scales are
been adequate. For example, Moos and Moos not excessively high; Moos and Moos (2009)
4 Family Environment Scale
reported that the largest correlation between them at nearly two standard deviations below the
scales was approximately 0.5 between cohesion mean (at a range of 30–35; M = 50, SD = 10
and conflict, whereas the correlation between for all scales). Similarly, they both reported rela-
active-recreational orientation and conflict is tively lower cohesion and higher conflict in the
near 0. This provides some justification for family when Jody’s brothers were visiting. Form
treating each scale as a measure of a distinct I revealed high congruence in Roger and Jody’s
construct. ratings of the ideal family environment, and the
Convergent validity of the FES has been tested therapist particularly emphasized their shared
in studies examining correlations between FES desires for more cohesion, control, and recrea-
scales and comparable subscales of other family tional activities. The therapist also reviewed
environment scales, including the Family Assess- answers to specific items on subscales in order to
ment Device (FAD) and the Family Adaptability draw attention to specific day-to-day concerns.
and Cohesion Evaluation Scales (FACES; Moos Based on this review, the family decided to hire
and Moos 2009). For example, cohesion as mea- part-time help to increase organization and free up
sured by the FES has been correlated with the time for mutually desired recreation. At a 3-month
cohesion subscales of the FAD and FACES. follow-up, organization was within one standard
The correlation between FES cohesion and deviation above of the norm. Both partners also
FACES cohesion has been reported by multiple rated family cohesion nearly one standard devia-
researchers as being in the range of r = 0.7–0.8 tion higher than prior to therapy; moreover, the
(Moos and Moos 2009). However, several of the clear communication about the disruption caused
subscales do not have counterparts in other family by Jody’s brothers allowed the family to focus on
environment measures. Some subscales, such as minimizing their influence.
moral-religious emphasis, have instead been
compared to indices of behavior (e.g., church
attendance, r = 0.6–0.7). Other subscales, like Cross-References
organization, have been compared to independent
observers’ ratings of variables like home cleanli- ▶ Beavers System Measures
ness (r = 0.5–0.6). ▶ FACES IV
▶ Family Assessment Device (FAD)
Moos, R. H., & Moos, B. S. (2009). Family environment Roosa, M. W., & Beals, J. (1990). Measurement issues in
scale manual: Development, applications and research family assessment: The case of the family environment
(4th ed.). Palo Alto: Mind Garden Inc. scale. Family Process, 29, 191–198. https://doi.org/10.
Nock, N. L., Ievers-Landis, C. E., Dajani, R., Knight, D., 1111/j.1545-5300.1990.00191.x
Rigda, A., Narasimhan, S., & Uli, N. (2016). Physical Sandhu, D. (2015). Predictive factors of health-risk behav-
activity self-efficacy and fitness: Family environment iours among male adolescents. Pakistan Journal of
relationship correlates and self-esteem as a mediator Psychological Research, 30, 1–19.
among adolescents who are overweight or obese. Sanford, K., Bingham, C. R., & Zucker, R. A. (1999).
Childhood Obesity, 12, 360–367. Validity issues with the family environment scale:
Pino, C., Simons, N., & Slawinowski, M. (1983). Devel- Psychometric resolution and research application with
opment and application of the children’s version of the alcoholic families. Psychological Assessment, 11,
family environment scale. Journal of Mental Imagery, 315–325. https://doi.org/10.1037/1040-3590.11.3.315
7, 75–82. Retrieved from http://www.journalofmenta
limagery.com/