Family-related life events

Insights from two Australian longitudinal studies
Research Report No. 22 – December 2012

Appendix F. Supplementary analyses, HILDA & LSAC: Change in personal wellbeing and the experience of life events

Multivariate analyses were carried out to examine the link between change in wellbeing measures and the experience of life events using fixed effect modelling.

Fixed effects models are useful for analysing the extent to which change in some characteristic is associated with a change in an outcome variable, such as a sense of wellbeing or distress. In the relevant HILDA and LSAC analyses, the fixed effects models focused on changes in outcomes between waves, with respect to characteristics of respondents or families that also changed across waves. In analysing change in this way, the effects of time-invariant characteristics that contribute to wellbeing or distress are swept out of the model. The fixed effects model can only include variables that change, and so a much more limited range of variables could be included in the specification. For example, the coefficients in Table F1 can be interpreted as indicating the degree to which psychological distress changes for those who experienced a life event, relative to any changes apparent for those who did not report the event.

Table F1: Multivariate analyses of changes in each wellbeing measure by socio-demographic characteristics, men, HILDA
  Life satisfaction Vitality Mental health Sense of social isolation Sense of social connection Overall wellbeing scale
Formed a live-in relationship –0.02   1.09   1.19   –0.22   –0.02   0.06  
Separated from spouse or long-term partner –0.28 *** –2.02   –3.72 *** 0.27 * –0.12   –0.2 ***
Pregnancy or birth/adoption of new child 0.21 ** –1.61   –0.46   0.02   0.01   0.01  
Serious personal injury/illness –0.25 *** –3.73 *** –2.96 *** 0.05   0.05   –0.14 ***
Serious injury/illness to family member 0.04   0.46   –0.06   –0.01   –0.08   0  
Death of spouse or child –0.05   0.13   –0.5   –0.51   0.25   0.13  
Death of close relative/family member –0.02   0.37   –0.03   0.13   0   –0.02  
Death of a close friend 0.01   –0.53   –0.41   0.02   0.05   –0.01  
Victim of physical violence –0.27 * –0.21   –2.66   0.37   –0.1   –0.16 *
Self or a family member detained in jail –0.11   –4.73 ** –0.84   0.11   –0.11   –0.13  
Retired from the workforce 0.04   0   0.33   –0.03   –0.07   0  
Fired or made redundant –0.04   –0.62   –1.44   –0.2   0.13   0  
Major worsening in finances –0.27 *** –1.6   –2.64 ** 0.16   –0.23 * –0.17 ***
Changed residence 0.13 ** 1.47 ** 0.51   –0.05   0.12 * 0.08 ***
Wave 10 –0.05 ** –1.02 *** –0.56 ** 0.21 *** –0.04   –0.02 *
Constant 7.93 *** 63.37 *** 76.92 *** 2.64 *** 7.25 *** 6.03 **
No. of observations 9,864 9,795 9,794 9,800 9,798 97,29

Note: * p < .05; ** p < .01; *** p < .001.

Source: HILDA Waves 9 & 10

Table F2: Multivariate analyses of changes in each wellbeing measure by socio-demographic characteristics, women, HILDA
  Life satisfaction Vitality Mental health Sense of social isolation Sense of social connection Overall wellbeing scale
Formed a live-in relationship 0.05   1.39   1.94 * –0.1   0.38 *** 0.13 **
Separated from spouse or long-term partner –0.05   –0.1   –0.97   0.25 * 0.28 * –0.02  
Pregnancy or birth/adoption of new child 0.04   –6.03 *** 0.17   –0.06   0.13   –0.04  
Serious personal injury/illness –0.14 ** –5.86 *** –4.41 *** 0.04   –0.09   –0.2 ***
Serious injury/illness to family member –0.04   –0.08   –0.99 * 0.06   0.08   –0.02  
Death of spouse or child –0.18   –4.72 * –7.32 *** –0.34   –0.23   –0.19 *
Death of close relative/family member 0.03   –0.81   –1.09 * 0.06   0.04   –0.02  
Death of a close friend 0   0.09   –0.08   0.03   0.00   0.00  
Victim of physical violence –0.11   –1.85   –6.63 *** 0.42 * 0.04   –0.2 *
Self or a family member detained in jail –0.01   0.65   –0.31   –0.13   –0.02   0.02  
Retired from the workforce –0.03   1.45   1.48   0.1   0.04   0.03  
Fired or made redundant –0.02   0.23   –1.48   0.13   –0.08   –0.04  
Major worsening in finances –0.41 *** –3.24 ** –3.99 *** 0.27 * –0.16   –0.25 ***
Changed residence 0.17 *** 1.15 * 1.2 * –0.1   0.02   0.09 ***
Wave 10 –0.03   –0.81 *** –0.59 ** 0.12 *** –0.04   0.00  
Constant 7.94 *** 59.3 *** 74.63 *** 2.48 *** 7.73 *** 0.03 **
No. of observations 11,260 11,198 11,198 11,172 11,171 11,100

Note: * p < .05; ** p < .01; *** p < .001.

Source: HILDA Waves 9 & 10

Table F3: Multivariate analyses of change in distress by experience of life events, LSAC
Characteristics as measured Wave 4 Fixed effects coefficient
Life events
Suffered a serious illness, injury or assault 0.45***
Had a serious illness, injury or assault happen to a close relative 0.28***
Had a parent, partner or child die 0.21*
Had a close family friend or another relative die 0.02
Was pregnant or had a baby –0.07
Had someone new move into the household 0.23*
Had a separation due to relationship or marital difficulties 0.42**
Broke off a steady romantic relationship 0.98***
Had a major financial crisis 0.86***
Lost job, but not from choice 0.12
Had something value lost or stolen 0.23*
Had someone in the household with an alcohol or drug problem 0.59***
Wave 2 (ref.)
Wave 3 0.39***
Wave 4 0.13**
Constant 2.78***
Sample size 22,250

Note: Change in Kessler 6 (scale 0 to 24, with higher value = more distressed). Only those life events that were available in three waves of data could be included. The life events that have been examined previously, that could not be included, are (a) started living with a new partner/spouse; (b) moved house; (c) lived in a drought-affected area; and (d) had home or local area affected by bushfire, flooding or severe storm. The wave indicator variables capture change that might have occurred more generally from wave to wave. * p < .05; ** p < .01; *** p < .001.

Source: LSAC Waves 2 to 4, B and K cohorts combined