Desktop version

Home arrow Education arrow Knowledge and Action


Sample Description and Drop-Out

The original German sample of N = 804 participants was disproportionately stratified by age and sex, resulting in almost equal subcategories of men and women (50 % each) and two age groups (51.2 % respondents aged 55-74 years and 48.8 % aged 75 years or older). The composition of the age groups changed from the first assessment in 1995 to the second in 2000 (61.3 % aged 55-74 years and 38.7 % aged 75 years or older) and even more dramatically from the second to the third assessment in 2005 (84.2 % aged 55-74 years and 15.8 % aged 75 years or older). By shifting the age group limit by 10 years and drawing on the participants’ actual age in 2005, we again obtained two groups of the same size (50 % each of respondents 65-74 years old and 75 years old or older).

The reasons for dropping out of the sample were documented in standardized protocols. Because of the long period of time, the most frequent reasons were the death of the former participants or a deterioration in their health (almost 20 % each). Other dropouts were due to refusals to continue participation and failure to locate or gain access to them. Logistic regression analysis based on data from the 2000 study indicated that age (OR = .94*),[1] education (OR = 1.59*), and the number of transport modes used in 2000 (OR = 1.27**) seemed to influence participation in 2005. Probability of participation increased with younger age, higher education, and greater variety of transport modes used. The level of education among the participants can be regarded as relatively high, with almost half of them having earned a standard or advanced degree.

As Table 15.1 shows, the average age of the 82 individuals who could be assessed over the 10-year interval was 75.2 years at T3, with 50 % of these participants belonging to the younger age group (65-74 years old) and 50 % to the higher age group (75 years old or older). Women and men were almost equally represented in the sample (48 % and 52 %, respectively). Most of the participants were married (66 %) and living in multiperson households (68 %). Approximately one in four (24 %) had lost their spouse. Satisfaction with their financial situation decreased on average from M = 7.7 in 1995 to M = 7.0 in 2005. Similarly, albeit at a lower level, subjective health was rated less positively 5 and 10 years after the first assessment (M = 6.9 and 6.7, respectively, compared to M = 7.3 in 1995).

  • [1] Odds Ratio is a way to quantify how strongly the presence of a variable A increases or reduces therisk that another variable B is present or absent. Risk is calibrated in this analysis such that 1.0means no change in the risk of B appearing when A is present. An OR of .94* means in our casethat being younger significantly reduced the risk that a participant in our study would drop out. *= significant at the .05 level; ** = significant at the .10 level (tentatively significant).
Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >

Related topics