Organizations, culture and behaviour: Opportunities, challenges and implications for crosscultural r

1
Organizations, culture and behaviourOpportunitie
s, challenges and implications for cross-cultural
research

• Ronald Fischer
• Centre for Applied Cross-Cultural Research
• Victoria University Wellington

2
Organizations, Culture and Behaviour (OCaB)
Project

• Multi-level study
• Exploration of relationships between variables at
  different levels
• Cross-level effects
• Units of investigation are nested within each
  other
• Compare the effect of organizational practices
  and cultural factors on employee behaviour
• Extra-role behaviour/Organizational Citizenship
  Behaviour

3
Challenges.

• Understanding organizational behaviour in context
• Single country/organization studies
• Two country/organizations comparisons
• Necessary
• studies including variables at multiple levels

4
Multilevel theories

• Exploration of relationships between variables at
  different levels
• Units of investigation are nested within each
  other

5
Example
6
Norms, values behaviour
Behavior
Norms Values (Descriptive, Injunctive)
Groups
Socialization (internalization)
Individuals
Values (Public, private)
Fischer, 2006, Personality and Social Psychology
Bulletin
7
Values as moderatorsFischer Smith (2004, JCCP
2006, APIR)
Decision-Maker
Is this fair ???
Use of allocation principles Equity Seniority
Values
8
Interaction between self-enhancement (high)
versus self-transcendence (low) values and
consideration of work performance (equity) on
justice
9
Important questions

• What is the relative effect of cultural values
  (compared to other non-cultural variables)?
• How do cultural values/norms affect (work)
  behavior?

10
Fischer, Ferreira, Assmar, Redford Harb, 2005,
IJCCM
Cultural Dimensions
Organizational Practices
11
Explaining cultural differences in organizational
reward allocationFischer et al. (2007, JCCP)
Conservation/ Hierarchy
Nation-Level
Unemployment rate
Organization Level
Organizational Culture (Economic/Egalitarian)
Sector
Individual Level
Need
Equity
Equality
12
Explaining national differences in organizational
commitmentFischer Mansell (submitted)
Hofstedes Power Distance
GLOBE In-Group Practices
GLOBE PD Practices
Hofstedes Individualism
GNI/GDP
Nation-Level
_
_
_
Organization Level
Method Effects
Occupation
Continuance Commitment
Affective Commitment
Normative Commitment
Individual Level
Cultural variables explained max. 10 of the
variance
13
Organizations, Culture and Behaviour (OCaB)
Project

• Compare the effect of organizational practices
  and cultural factors on employee attitudes and
  behaviour
• Affective Commitment
• Extra-role behaviour/Organizational Citizenship
  Behaviour
• Perceived organizational justice

14
Organizational Practices
Nation-Level
Power Distance
Collectivism
Fatalism
Paternalism

Organization- Level
_
_
Employee Orientation
Innovation
Formalization
Individual Level
_
_
Helping
Voice
15
Culture level dimensionDescriptive Norms
(similar to GLOBE)

• Power distance
• Acceptance of hierarchy
• Social distance
• Egalitarian relations
• Paternalism
• Fatalism
• Control
• Inactivity
• Uncertainty avoidance
• Structured lives
• Rule obedience
• Individualism Collectivism (Triandis, 1995)

16
Individualism-Collectivism

• Dominant paradigm for studying cultural
  differences in psychology
• Confusion about meaning definitions
• Oyserman et al. (2002) 7 Ind and 8 Col domains
• Single-bipolar or independent dimension
  (Hofstede, 1980, Gelfand et al., 1996, Triandis
  et al.)
• Confusion about measurement
• Self-ratings on Likert-type scales
• Neglect of relevant in-group (exception Hui,
  1988)
• Reference-group effects (Heine et al., 2001)
• Abstract meanings, cultural anchor missing

17
Defining attributes of Individualism
Collectivism (Triandis, 1995)

• Definition of self
• Structure of goals
• Norms versus attitudes
• Relatedness versus rationality

18
New scale

• Four defining attributes
• Single dimension
• Specifying important in-groups
• Appropriate referent for cultural comparison
  importance within ones culture
• referent-shift consensus model (Chan, 1998)
• Provide meaningful anchors examples of
  individualistic and collectivistic behaviour
• similar to semantic differential type scales

19
Constructing items

• Team process (international collaboration)
• Self (Gudykunst et al., 1994 Singelis, 1994)
• Structure of goals (Yamaguchi, 1994)
• Norms versus attitudes (all new)
• Rationality versus relatedness (Clark et al.,
  1987 Gudykunst et al., 1994 Singelis, 1994)
• Three pilot studies

20
Scale development

• Study 1
• Brazilian (N 517), Taiwanese (N 178), NZ (N
  210) and NZ international students (N 89)
• Cross-validation Brazilian (N 89) NZ (N 85)
  and Taiwanese (N 62) employees
• Between 3 and 4 factors in each sample (Scree
  test)
• Structure of goal items (Yamaguchi, 1994) did not
  clearly load
• Internal consistencies
• Self (a range .52 - .77)
• Norms (a range .80 - .84)
• Rat. Vs. Relat. (a range .78 - .89)

21
Study 2

• Objectives
• Construct new items for goal attributes
• Validate with different samples
• Confirmatory factor analysis
• Between culture variability
• Further evidence of convergent validity
• Vertical and Horizontal IndCol

22
Study 2

• 1,680 social science and psychology students
• Argentina (N 189),
• Brazil (N 255),
• Germany (N 108),
• India (N 152),
• Lebanon (N150),
• New Zealand (N 158),
• Peru (N 133),
• Saudi Arabia (N107),
• Taiwan (N 137),
• United Kingdom (N 42) and
• United States (N 183)
• Dimensionality and validity at an individual and
  cultural level

23
Confirmation of the Structure

• Overall model (pooled within-group matrix)
• ?2(203) 1195.91, TLI .95, CFI .96, RMSEA
  .058 (with revised model, item 17 loading on
  Self- Factor)
• Multigroup analysis
• ?2(2876) 6931.40, TLI .93, CFI .93, RMSEA
  .076
• Means and Covariance Structure Analysis (MACS)
• ?2(2733) 5079.37, TLI .92, CFI .92, RMSEA
  .081
• No strong evidence of uniform or non-uniform item
  bias (response styles unlikely)
• Multilevel equivalence
• ?2(424) 1437.87, TLI .93, CFI .94, RMSEA
  .057

24
Agreement within cultures

• The average agreement across all items and
  cultural samples was relatively low (.43).
• Mean for self (.39), attitudes versus norms
  (.43), relational versus rational concerns (.41),
  goal structure (.50)
• High agreement UK (.59), New Zealand (.56),
  Germany (.53)
• Low agreement Lebanese (.31) and Indian samples
  (.34)

25
Convergent validity

• Individual level
• Significant correlations with vertical-horizontal
  individualism-collectivism (Triandis Gelfand,
  1998) especially HI VC
• At culture level
• Self x Hofstede IDV r .65
• Self x GLOBE In-Group Col Practices r -.62
• Self x GLOBE Institutional Col values r -.67
• No other significant correlations
• No significant correlation with PD (Hofstede,
  GLOBE)

26
Study 3

• Study with 161 NZ students
• New scale explained additional variance beyond
  established scales (e.g., Triandis Gelfands
  Vertical-Horizontal Scale) in conformity oriented
  and altruistic behaviour
• But normative IC does not explain variance in
  self-oriented behaviour

27
Implications for cultural measurement

• When interested in cultural comparison
• Treat participants as informants (Glick, 1985)
  use Referent-shift consensus models (Chan, 1998,
  see Klein et al., 1994)
• Referent Culture in general
• Specify in-group of interest
• Provide anchoring points
• When interested in relating personal cultural
  attributes to other person-level variables
• Use individual-level referents
• Specify relevant in-group provide anchoring
  points

28
Discussion

• Construct validity at individual and cultural
  level
• Good overlap with previous individual level
  measures
• Consistent relationship between self scale and
  previous individualism-collectivism scales
• Important theoretical aspects have been missing
  in previous research!

29
Typology of Culture InventoriesAshkanasy,
Broadfoot Falkus 2001
Level 1 Artefacts Level 1 Patterns of Behaviour
Types
Profiles
Behavior Fit
Effective Behaviour
Behavioral Patterns
Level 2 Beliefs and Values
Types
Profiles
Value Fit
Descriptive Values
Effectiveness Values
Level 3 Basic Assumptions
30
Organizational practices

• Practices are observable and visible
  manifestations of organizational culture
  (compared with underlying values, norms and
  beliefs) such as symbols, heroes, rituals,
  alternative labels include conventions, customs,
  habits, mores, traditions and usages (Hofstede,
  Neuijen, Ohayv, Sanders, 1990)

31
Fischer et al. 2005 IJCCM
32
Item development

• Team process (Simultaneously in English,
  Portuguese Spanish)
• Translation backtranslation into Turkish
• English version in Malaysia
• 71 items derived from the literature or newly
  developed
• Focus on specific practices
• Criteria
• Related to a general organizational culture
  dimension
• Relevance for one of our outcome variables

33
Item format

• Focus on observable and visible practices
• Examples
• Whenever an employee is ill, managers ask about
  him/her with interest
• Targets to be attained by employees are clearly
  specified
• There is a lot of investments in new products in
  this organization
• Instructions
• Below you see a number of statements about work
  practices and behaviours. Please indicate how
  frequently each of these situations occurs in the
  organization in which you work.
• 7 point Likert scale (never to always)

34
Organizational Pilot Study

• 1243 employees from various organizations in
• Argentina (N 215),
• Brazil (N 279),
• Malaysia (N81),
• New Zealand (N 310),
• Turkey (N 200), and
• US (N 158)

35
Demographics

• Equal number of males and females in all samples
  (about 50)
• Majority employed in private sector (62), but
  less in NZ US
• 1/3 middle management 1/3 process/clerical
  workers about 1/10 senior management
• Majority employed in service industry (60),
  followed by primary/manufacturing (30)

36
Factor structure

• Employee orientation (work organization, task
  versus employee orientation dimensions)
• Innovation (Innovation dimension)
• Formalization and Bureaucratic practices (Control
  dimension)
• Mean congruence (Tuckers Phi) excellent (.97)
  structure clearly replicated in all samples

37
Reliability

• Employee orientation (10 items)
• Range from .92 to .95
• Innovation (7 items)
• Range from .74 to .89
• Formalization Bureaucratic practices (7 items)
• Range from .78 to .87

38
Effects on work behaviour

• Hofstede (1980) scores at the nation-level
• Organizational Practices ERB at the
  individual-level
• Cross-Level Operator Analysis (equivalent to
  Hierarchical Linear Modeling)

39
Nation-Level
Economic Growth
Individualism
Individual- Level
Formalization
Employee orientation
Innovation
Helping
Organizational Practices effects independent
of Cultural variables
40
Nation-Level
Individualism
Economic Growth
Individual Level
Employee orientation
Innovation
_
Formalization
Voice
41
Study 2

• 1541 employees from 50 organizations in NZ
• 471 from 19 organizations in Lebanon

42
Aggregation statistics

• NZ
• Aggregate reliability
• ICC(1) between .16 (employee orientation) to .30
  (formalization)
• ICC(1) for voice (.06) and helping (.05)
• Agreement
• Mean awg between .53 (employee orientation) and
  .55 (innovation)
• Mean awg for voice helping (.51)

43

• Lebanon
• Aggregate reliability
• ICC(1) between .03 (formalization) to .26
  (innovation)
• ICC(1) for voice (.26) and helping (.34)
• Agreement
• Mean awg between .24 (employee orientation) and
  .30 (innovation)
• Mean awg for voice (.32) helping (.35)

44
HLM Analysis

• Level 2 Organizational Practices (employee
  orientation, innovation, formalization)
• Level 1 Individual level self-reported
  extra-role behaviour (helping, voice)

45
NZ significant effects
Organization Level
Employee Orientation
Innovation
Formalization
.21
Individual Level
Helping
Voice
Correlation between employee orientation voice
at organization level .27, p .05
46
Lebanon significant effects
Organization Level
Employee Orientation
Innovation
Formalization
-.64
-.53 p .07
1.52
1.81
Individual Level
Helping
Voice
47
Discussion

• Effects of organizational practices are moderated
  by nation-level norms values
• Pro-innovation bias in Western societies might
  not generalize
• Formalization can have positive effects in some
  contexts
• Level of analysis issues potential fallacies
  (e.g., significant correlation at aggregate
  level, but not in HLM agreement reliability
  issues)

48
Research in progress

• Aim 20 organizations in 20 countries 3 level
  structure (nation, organization, individual)
• Investigate separate effects of nation-level
  values norms and organizational practices
• To date
• Organizational practice effects are not universal
• Nation-level values moderate some effects (e.g.,
  INDCOL), but significant cross-national variation
  is not accounted for by existing nation-level
  indicators

49
Data collected

• NZ
• Lebanon
• Brazil
• Malaysia
• Taiwan
• Philippines

• Egypt
• Saudi Arabia
• UAE
• UK
• Indonesia
• (Singapore)
• (Belgium)