Evaluating the inter-respondent (consumer vs. staff) reliability and construct validity (SIS vs. Vineland) of the Supports Intensity Scale on a Dutch sample.
The Dutch SIS is valid, but staff and consumer scores won't match exactly—pick the rater that fits your purpose.
01Research in Context
What this study did
The team gave the Supports Intensity Scale (SIS) to Dutch adults with intellectual disability. They also asked staff to fill out the same form.
Next they compared the two sets of scores and checked them against the Vineland Adaptive Behavior Scales.
What they found
Staff and consumer ratings moved in the same direction, but the numbers were not identical. When SIS scores went up, Vineland scores went down, just as expected.
This pattern shows the Dutch SIS works the way it should.
How this fits with other research
Schaaf et al. (2015) later asked the same question in Australia. They found that support-needs scores alone predicted funding levels, while adding adaptive behavior added no extra power. This extends the Dutch finding: if you want to decide dollars, reach for a support-needs tool, not an adaptive one.
Davis et al. (2009) tested the Rosenberg Self-Esteem Scale in the same year and got poor validity for adults with ID. The Dutch SIS succeeded where the RSES failed, showing that not all self-report tools survive translation.
Rasing et al. (1992) wrote the early roadmap: translate, back-translate, pilot, then check stats. Bigby et al. (2009) followed those exact steps, proving the old guide still works.
Why it matters
When you pick a plan, decide whose voice you need. If you want the person's own view, use the consumer SIS. If you want the team's view, use the staff SIS. Either way, expect the numbers to be close but not twins. Use the SIS alone for funding decisions; adding Vineland or ICAP won't boost accuracy.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Try collecting both staff and consumer SIS on one client and note where the scores differ.
02At a glance
03Original abstract
BACKGROUND: Despite various reliability studies on the Supports Intensity Scale (SIS), to date there has not been an evaluation of the reliability of client vs. staff judgments. Such determination is important, given the increasing consumer-driven approach to services. Additionally, there has not been an evaluation of the instrument's construct validity on a non-English speaking sample. This is important as the SIS is currently translated into 13 languages. METHOD: Data were collected in two different samples, using the Dutch translation of the SIS and the Vineland-Z. RESULTS: There was a significant correlation between ratings of staff and consumers on the SIS; however, the relationship between the mean scores of consumer and staff responses indicated significant differences in staff and consumer scores. All correlations between the Vineland-Z domains and the SIS subscales were significant and negative, ranging from -0.37 to -0.89. CONCLUSIONS: Analyses of the inter-respondent reliability suggest that one needs to consider the source of information regarding needed supports carefully. The significant negative correlations between SIS and Vineland-Z reflect that the SIS is measuring a different construct (needed support) than the Vineland-Z (adaptive behaviour). The results of the two studies provide additional support for the etic (universal) properties of the SIS, as both hypotheses were confirmed. In conclusion, SIS users are provided with a wealth of information that can be used for multiple purposes.
Journal of intellectual disability research : JIDR, 2009 · doi:10.1111/j.1365-2788.2008.01149.x