Corrigendum to “Knowing, planning for and fearing death: Do adults with intellectual disability and disability staff differ?” [Res. Dev. Disabil. 49–50 (2016) 47–59](S089142221530010X)(10.1016/j.ridd.2015.11.016)

The authors regret that our original article contained errors. The errors were due to (a) a small number of data entry errors (affecting Table 3 and related analyses reported in Section 3.3, and Table 4 and related statistical analyses reported in Section 3.4), and (b) transposition errors (affecting values of F in Table 4). These minor corrections do not change our findings, our interpretation of these findings, or our conclusions. The authors apologise for any inconvenience. The corrected tables and text are shown below. For the sake of clarity, the complete corrected tables are shown, as is the text from the entire corrected subsection in which analyses are reported. 3.3. Concept of death CODQ component raw scores and total raw scores by group are shown in Table 3. One participant with ID chose not to be assessed on the CODQ so n = 38 for the ID group in Table 3. For CODQ scores overall, based on the multivariate analysis, participants with ID had a significantly poorer understanding of death than disability staff, F (5, 70) = 2.49, p = 0.039. This difference was contributed to by significant univariate group differences for the Non-functionality and Inevitability component scores (Table 3). The covariates sex (F (5, 70) = 0.99, p = 0.432) and age (F (5, 70) = 0.71, p = 0.618) were not significant for the overall multivariate comparison or for any of the univariate analyses of the five CODQ component scores. 3.3.1. Analyses among participants with ID We compared the CODQ total scores for participants with mild ID (M = 25.08, SD = 2.58, range 14–26) versus moderate ID (M = 22.08, SD = 4.06, range 16–26) and found a significant difference, t = 2.34, p = 0.017 (1-tailed). The effect size was large, d = 0.88. There were no significant differences (t = 0.12, p = 0.91, 2-tailed) between men with ID (M = 24.21, SD = 3.34, range 16–26) and women with ID (M = 24.07, SD = 3.52, range 14–26). Among the ID group, the percentage with a complete understanding of each CODQ component (i.e., gained the maximum score for that component) was 92.1% for Causality, 97.4% for Finality, 86.8% for Non-functionality, 89.5% for Universality, 78.9% for Inevitability, and 68.4% for CODQ Total scores. Comparison using within-subjects simple contrasts among the four CODQ components with a scale maximum of six (i.e., excluding Causality), yielded a significant difference in the expected direction with the Inevitability component scores being significantly lower than the Finality component, F(1, 37) = 5.03, p = 0.031. The differences between Inevitability component scores and the other two CODQ components (Non-functionality, Universality) were in the same direction but did not quite attain significance (p < 0.10 for both). 3.4. End-of-Life Planning Scale (EOLPS) EOLPS subscale raw scores and total raw scores by group are shown in Table 4. EOLPS scores are presented and analyzed in two ways, first by subscale (the total of the three item scores for each of the seven subscales such as Bequeathing), and by question type (the total of all seven presentations of each question, such as Q1 “Who decides if…”). 3.4.1. Subscales Based on the multivariate analysis of the seven subscales, for EOLPS scores overall, participants with ID had significantly lower scores of end-of-life planning than disability staff, F (7, 63) = 15.80, p = < 0.001. This difference was contributed to by significant univariate group differences for the all seven subscale scores (Table 4). The covariates sex (F (7, 63) = 0.77, p = 0.612) and age (F (7, 63) = 1.80, p = 0.104) were not significant for the overall multivariate comparison. 3.4.2. Questions Likewise, when the data were aggregated by the three types of question, multivariate analysis showed that participants with ID had significantly lower EOLPS scores than disability staff, F (3, 67) = 40.46, p = < 0.001. The covariates (gender, age) were not significantly related to EOLPS scores under either multivariate or univariate analyses. Univariate analysis showed that participants with ID had significantly lower scores on each of the three questions (last 3 rows of Table 4), especially for question 1. Indeed, an average of 62.5% of participants with ID had a Q1 score of zero (others decide) on these seven items (range 50–78% across items). By contrast, only 1.4% of staff responses to question 1 stated that someone else decides. Effect sizes were large for all comparisons. 3.4.3. Analyses among participants with ID Participants with mild ID (M = 28.64, SD = 9.07, range 8–42, n = 22) versus moderate ID (M = 14.00, SD = 8.82, range 1–30, n = 11) differed significantly (t = 4.41, p < 0.001, 1-tailed) on EOLPS total scores. There was a very large effect size of d = 1.30. EOLPS total scores for men with ID (M = 22.76, SD = 11.38, range 1–42, n = 21) and women with ID (M = 25.50, SD = 11.41, range 8–40, n = 12) did not differ significantly (t = −0.66, p = 0.511, 2-tailed). Among the participants with ID, the lowest mean score was for the preferred place of care subscale and the highest for the organ donation subscale. Comparisons using within-subjects simple contrasts among the seven EOLPS subscales revealed significant differences between Preferred place of care and all other subscales except advance care planning and bequeathing. Likewise, there was a highly significant difference in scores across the three EOLPS questions with the total for Question 1 being the lowest, and Question 2 being the highest. Within-subjects simple contrasts showed that relative to Question 1, scores were significantly higher for Question 2, F (1, 32) = 59.71, p = <0.001, and Question 3, F (1, 32) = 7.55, p = 0.010.