WHAT MAKES PEOPLE HAPPY

1
WHAT MAKES PEOPLE HAPPY?
Social Support Emotional Intelligence as
Predictors of Subjective Wellbeing Emma
Gallagher Dianne Vella-Brodrick
2
What we already know

• Happiness is operationalised by Subjective
  wellbeing (SWB).
• SWB studies highlight factors that foster optimal
  psychological functioning.
• Three components of SWB Positive Affect (PA),
  Negative Affect (NA), Satisfaction with Life
  (SWL).

3
Sociodemographic variables personality factors

• Sociodemographic variables 8-20 (Argyle, 2001
  Diener et al., 1999).
• Personality factors consistently large proportion
  of variance.
• e.g. 34 of unique variance in a general
  community sample (Gannon Ranzijn, 2005).

4
Social support

• Social support (SS) relates positively to SWB.
• Some believe SS is necessary for SWB (Baumeister
  Leary, 1995 Diener Oishi, 2005, Diener
  Seligman, 2002).
• SS thought to promote well-being through the
  provision of resources from one person to another
  which influence emotions, cognitions, and
  behaviours that in turn help people cope and
  enjoy life.

5
Social support considerations

• Need to measure SS and all three components of
  SWB.
• Source of SS (significant other, family, friends)
  over type of support (material aid).
• Perception of SS over number of supports and/or
  receipt of support.
• SWB source perception of SS
  sociodemographic variables personality factors

6
Emotional intelligence

• Linked to both SWB and SS (Bar-On, 2005 Salovey,
  Bedell, Detweiler, Mayer, 1999).
• Relatively ignored in SWB research.
• Eminent SWB researchers have suggested EI is
  worth investigating for variance in SWB.
• Investigations of SWB/EI relationship are
  relatively new.

7
Emotional intelligence

• EI was labelled and modelled by Salovery and
  Mayer (1990).
• Popularised by Goleman (1995).
• Broadly defined as the cognitive ability to
  perceive, manage, and regulate emotions within
  ones self, and others, in ways that maximise
  positive cognitive and behavioural outcomes that
  result in more beneficial life outcomes (Bar-On,
  2005 Mayer Salovey, 1997 Salovey et al.,
  1999 Salovey Mayer, 1990).

8
Emotional intelligence

• Emerging evidence that EI can be taught, opposed
  to other SWB predictors such as personality
  (Emmerling Goleman, 2003 Reshmi, 2006 Slaski
  Cartwright, 2002 Stys Brown, 2004).
• Utility of EI complicated as more than one model
  has been proposed.
• Conjecture regarding the discriminant validity of
  EI now sufficient evidence showing discriminant
  validity (Bar-On, 2005 Brackett Mayer, 2003
  Ciarrochi, Chan Bajgar, 2001 Ciarrochi, Chan,
  Caputi, 2000 Ciarrochi, Deane Anderson, 2002
  Gannon Ranzijn, 2005 Lopes et al, 2004
  Schette et al., 1998 Tett, Fox Wang, 2005).

9
Emotional intelligence

• EI has been shown to relate positively to SWL and
  PA.
• SWL influenced by how clearly people understand
  emotion.
• Controlling for well-known predictors of SWB
  provide clearer results for EI.
• Research need use wellrespected SWB measures
  measure 3 x components of SWB and control for
  well-established predictors of SWB.

10
SWB, SS EI

• Underlying processes of SS and EI seem similar.
• The moderator question is
• Does the relationship between SS and SWB depend
  on an individuals level of EI?

11
Hypotheses

• That where sociodemographic variables and
  personality factors are controlled, SS from
  Significant Other, Family and Friends would
  significantly predict SWL, PA and NA. It was also
  hypothesised that where sociodemographic
  variables and personality factors are controlled,
  EI would significantly predict SWL, PA and NA. It
  was further hypothesised that EI would
  significantly influence the relationship between
  SS and SWB as measured by sources of support and
  SWL, PA and NA respectively, where the
  interaction effects between SS and EI would add
  significant variance to the prediction of SWB
  beyond main effects.

12
Participants

• 267 adults 196 females, 71 males.
• general population who volunteered after
  reading explanatory statement.
• Age 18-80 (M41.52 years, SD14.28).
• 52.8 in relationships.
• 72.6 were tertiary educated.
• Income range 80-2699 (gross) per week, modal
  income 1000 per week.

13
Measures
Response Type all self report
Plus socio-dem questions
Alpha (Original) This Study
Item example
of Items
5 point Likert Scalevery slightly or not at all
to, extremely
Positive NegativeAffect Schedule 2 subscales
(PANAS)
20 10 e/s
Indicate how you feel eg. interested Ref Watson
et al. (1988)
PA, (.88) .88 NA, (.87) .89
7 point Likert Scale Strongly Disagree to,
Strongly Agree
I am satisfied with life Ref Diener et al. (1985)
Satisfaction With Life Scale (SWLS)
5
(.87) .89
Multidimensional Scale of Perceived Social
Support - 3 subscales (MSPSS)
12 4 e/s
There is a special personwhen I am in need Ref
Zimet et al. (1988)
7 point Likert Scale Very Strongly Disagree to,
Very Strongly Agree
SO (.91) .95 FAM (.87) .94 FRI (.85) .94
5 point Likert Scale Strongly Disagree to,
Strongly Agree
Schutte Emotional Intelligence Scale (EIS)
33
I have control over my emotions Ref Schutte et
al. (1998)
(.90) .90
International Personality Item Pool 5 subscales
(IPIP)
E (.87) .88 A (.82) .75 C (.79)
.84 ES (.86) .93 II (.84) .76
6 point Likert Scale Very inaccurate to, Very
accurate
50 10 e/s
Am the life of the party Ref Goldberg (1999)
Plus Ballards short social desirability
14
Procedure

• Questionnaire kits with posters.
• Public places.
• Explanatory statement.
• Anonymity and confidentiality.
• 20 minutes.
• Australia Post or returns box.

15
Analyses

• Missing values and violations of the assumptions
  of multivariate analysis
• 8 univariate outliers truncated
• Mean replacement where missing values were less
  than 5
• N 267-234 on pairwise analyses
• No order effects
• Main and interaction effects
• Hierarchical Multiple Regression in SPSS

16
HMR steps
17
Satisfaction with Life

• Model as a whole predicted 44.1 (40.7 adjusted)
  R.66, F(13, 216)13.09, plt.001
• 1 Sociodem Personality 34
• 2 SS 2.9
• 3 EI 2.6
• 4 SS x EI (interaction) 2.7
• - EI, and SSso x EI were significant at the
  last step
• significant at .05

18
Positive Affect

• Model as a whole predicted 47.7 (44.9
  adjusted) R.69, F(13, 239)16.78, plt.001
• 1 Sociodem, Personality Sdes 41.3
• 2 SS (2)
• 3 EI 3.3
• 4 SS x EI (interaction) (1)
• - EI significant in last step
• significant at .05

19
Negative Affect

• Model as a whole predicted 50.3 (47.6 adjusted)
  R.71, F(13, 239)18.60, plt.001
• 1 Sociodem, Personality Sdes 45
• 2 SS 2.1
• 3 EI (1)
• 4 SS x EI (interaction) (1)
• - SS Friends x EI was significant at the
  last step
• significant at .05

20
Interactions

• SWL SS Significant Other x EI
• NA SS Friends x EI
• Interactions indicate that the relationship
  between SS and SWB is dependent on the level of
  EI reported.

21
SWLSS Significant Other x EI
EI low, b 0.27 t 3.18 p .0008 EI high,
b -0.05 t -0.56 p 0.2896
22
NASS Friends x EI
EI low, b -0.22 t -2.79 p .0028 EI
high, b .07 t .88 p 0.1896
23
Discussion

• The results partially support the hypotheses.
• SS predicted SWL and NA, but not PA.
• EI predicted SWL and PA, but not NA.
• 2/9 interaction terms significant.
• Each model as a whole was significant.

24
Discussion

• Small amount of variance accounted for by SS,
  inconsistent with previous research.
• Possible explanation is stringent controls
  however can be seen to provide clearer results
  for unique value of SS.
• Differential weights of SS sources across DVs
  important consideration.

25
Discussion

• Adds support to the discriminant validity of EI.
• Predictive value of EI in SWB exciting, as EI
  thought to be subject to change.
• Significant interaction terms support suggestions
  that EI is important to SWB and SS.

26
Discussion

• SS x EI interaction re SWB not previously
  published.
• Where EI is high the level of perceived SS is
  inconsequential to the level of SWB reported.
  However, where EI is low the level of perceived
  SS and level of EI interact to produce a
  differential level of SWB, with higher reported
  SS having a relationship with higher SWB .

27
Discussion

• SS may not be necessary for everyone,
  specifically those with high EI.
• EI training could be considered as an alternative
  to formal SS.
• Mindful that these are new results with some
  limitations gender and education split.

28
Discussion

• Valuable insight into SS and EI as predictors of
  SWB.
• Addressed limitations of earlier studies.
• Shows that continuous investigations into
  apparently robust relationships are warranted.

29
Discussion

• Professionals concerned with SWB need to consider
  main and interactions effects of known variables.
• This study shows that the relationship between SS
  and EI on SWB goes beyond main effects, and that
  it is important to explore the three components
  of SWB.

30
Thank you