Two sides of the same coin: Are the Dual Types of Passion Types in the Sense of Distinct Subgroups of Individuals?

Two sides of the same coin:

Are the Dual ‘Types’ of Passion Types in the Sense of Dis7nct Subgroups of Individuals? Manuscript accepted for publica7on at the Journal of Person-‐Oriented Research

Julia Moeller1, Melanie Keiner2, & Robert Grassinger3

1 = Yale Center for Emo7onal Intelligence, US; 2 = University of Erfurt, Germany; 3 = University of Augsburg, Germany.

“Most passionate individuals strongly endorse the passionate

items from both subscales,

making it impossible to distinguish “pure” harmoniously

passionate people (those who did not also endorse items from

the obsessive subscale) from “pure” obsessively passion people

without losing the majority of the sample to a mixed group.”

(Mageau et al., 2009, p. 613 )

Main topics of this talk

1. Do individuals experience either mainly harmonious, or mainly obsessive passion, or are both ‘types’ of passion experienced together?

2. Why z-‐scores are misleading in the examinaDon of profiles.

DefiniDon of Passion

Passion is defined as “a strong inclinaDon toward an acDvity that people like, that they find important, and in which they invest Dme and energy” (Vallerand et al., 2003, p. 755).

Types of Passion “People with an harmonious passion are able to decide to terminate the relaDonship with the acDvity if they decide it has become a negaDve factor in their life.” (Vallerand et al., 2003, p. 758).

“People with an obsessive passion can thus find themselves in the posiDon of experiencing an uncontrollable urge to partake in the acDvity they view as important and enjoyable.” (Vallerand, 2012, p. 3).

HP and OP are correlated with each other.

What is a type?

•  Types describe subgroups of individuals with homogeneous profiles

•  The classic methods to classify individuals into groups with similar profiles is cluster analyses

Cluster analyses on Passion Cluster analyses are very rare in the research on passion.

Wang et al. (2008) and Wang & Yang (2008) found that individuals experienced aligned HP and OP, meaning these variables were either both high, or both low, or both moderate, within individuals, while HP tended to be generally higher than OP.

OperaDonalizing types of passion (2)

1.  HP and OP are standardized across all individuals.

2.  Non-‐passionate individuals are separated (non-‐passionate = passion criteria mean score below the scale midpoint).

3.  The remaining passionate individuals are classified into the ‘mainly harmonious group’ (zHP > zOP) and the ‘mainly obsessive group’ (zOP > zHP).

-‐2.40

-‐2.00

-‐1.60

-‐1.20

-‐0.80

-‐0.40

0.00

0.40

non-‐passionate group

HP group OP group

Study 1 z scores

z HP z OP

Mageau et al. (2009),

Philippe, Vallerand, & Lavigne (2009),

Vallerand and Houlfort (2003).

Vallerand and colleagues used another classificaDon procedure, the z-‐score-‐based classificaDon:

Research QuesDons 1.  Do all passionate individuals experience either

predominantly harmonious or predominantly obsessive passion, or are there groups with generally high, or low, levels in both HP and OP?

2.  Does the grouping procedure based on z-‐standardized scores adequately differenDate between individuals with higher HP than OP (harmonious individuals) and individuals with higher OP than HP (obsessive individuals)?

3.  Do alternaDve grouping procedures such as grouping based on raw scores and cluster analysis produce the same groups and point to similar conclusions?

Study Sample & assessment N Instruments

Study 1 German undergraduate students (UniversiDes of Erfurt

and Augsburg),

471 dual model passion scale a,

com.pass scale b,

Affect scalec

Study 2 Brazilian adolescents in two private schools in Rio de

Janeiro


com.pass scale b,

sport commitment scaled

Study 3 German adolescent soccer player, online survey and

paper-‐&-‐pencil quesDonnaires

278 dual model passion scale,

com.pass scale a,

sport commitment scaled,

BASe,

Affect scale: PANASf

Study 4 German leisure Dancers (149), marDal arts pracDDoners

(95)

244 dual model passion scalea,

com.pass scale b,

BASe,

Dependency measure EAIg

Samples and Measures Study Sample & assessment N Instruments

Study 1 German undergraduate students

(UniversiDes of Erfurt and Augsburg),


com.pass scale b,

Affect scalec

Study 2 Brazilian adolescents in two private

schools in Rio de Janeiro


com.pass scale b,

sport commitment scaled

Study 3 German adolescent soccer player, online

survey and paper-‐&-‐pencil

quesDonnaires

278 dual model passion scale,

com.pass scale a,

sport commitment scaled,

BASe,

Affect scale: PANASf

Study 4 German leisure Dancers (149), marDal

arts pracDDoners (95)

244 dual model passion scalea,

com.pass scale b,

BASe,

Dependency measure EAIg

Passion 1: Dual Model Passion Scale (Vallerand et al., 2003);

Passion 2: com.pass scale (Moeller & Grassinger, 2014b);

PANAS (PosiDve NegaDve Affect Schedule) (Krohne, Egloff, Kohlmann, & Tausch, 1996);

Affect Scale (Kessler & Staudinger, 2009);

Sport Commitment Scale (Sousa et al., 2008);

Behavioral AcDvaDon, subscale of the BIS/BAS scale (Carver & White, 1994; Strobel, Beauducel, Debener, & Brocke, 2001);

Exercise AddicDon Inventory (Terry, Szabo, & Griffiths, 2004).

Measures

Research QuesDon 1

Do all passionate individuals experience either predominantly harmonious or predominantly obsessive passion,

or are there groups with generally aligned high, or aligned low, levels in both HP and OP?

Aligned HP and OP: Scaqerplots of raw scores

Obsessive individuals: 1




HP and OP are aligned within individuals, i.e. either both high, or both low, or both moderate, within individuals.

There is no clear disDncDon between high and low passion. Instead, both HP and OP increase conDnuously.

Even though HP and OP are aligned, HP tends to be higher than OP for most individuals.

Hardly any ‘mainly obsessive’ individuals (OP > HP) are found.

Moeller, J., Keiner, M., & Grassinger, R. (accepted). Two sides of the same coin: Are the dual ‘types‘ of passion types in the sense of disDnct subgroups? Journal for Person-‐Oriented Research.

Research QuesDon 2

Does the grouping procedure based on z-‐standardized scores adequately differenDate between individuals with higher HP

than OP (harmonious individuals) and individuals with higher OP than HP (obsessive individuals)?

MisinterpretaFons: Grouping via raw scores (above) versus grouping via z-‐scores (below)

Above: Looking at the raw scores, we see that both the ‘harmonious’ and the ‘obsessive’ group have higher scores in harmonious than in obsessive passion!

Below: Looking at the z-‐scores of the same (!) groups, it looks like there is one group with high harmonious scores, and another groups with high obsessive scores.


Puzzling z-‐scores SomeDmes z-‐scores suggest that response A was higher than response B, even though it was the other way around when the person answered to the original response scale!

z-‐scores can also suggest that an answer was “high”, even though this answer negated the item statement on the original response scale (=was a raDng below the scale midpoint on a scale from 1 = don’t agree at all to 7 = totally agree).

Why is that? Does this only occur in passion research?

Why misinterpretaFons? DistribuDons of Variables

Research QuesDon 3

Do alternaDve grouping procedures such as grouping based on raw scores and cluster analysis produce the same groups and

point to similar conclusions?

Percentages of misleadingly classified individuals

Study 1 Study 2 Study 3 Study 4

HPraw > Opraw

, but labelled ‘obsessive’

with the z-‐score procedure

38.8% 21.6% 32.6% 40.2%

OPraw > Hpraw

, but labelled ‘harmonious’

with the z-‐score procedure

0.0% 0.0% 0.0% 0.0%

Sizes of groups of non-‐passionate, harmonious (HP > OP), obsessive (OP > HP) and equally harmonious and obsessive (HP = OP) individuals

Study 1 Study 2 Study 3 Study 4

z-‐

score

Raw

score

z-‐

score

Raw

score

z-‐score Raw

score

z-‐

score

Raw

score

Nonpassionate 1.5% 1.5% 5.9% 5.9% 2.7% 2.7% 3.7% 3.7%

HP > OP 53.9% 93.2% 46.2% 67.8% 44.7% 77.3% 49.2% 89.3%

OP > HP 44.7% 5.3% 48.0% 20.1% 52.7% 9.5% 47.1% 5.3%

HP = OP 0% 0% 0% 6.2% 0% 10.6% 0% 1.6%

Results cluster analysis Individuals are either high, or low, in both HP and OP.

High passion comes with high commitment, disposiDonal approach moDvaDon, and dependency.


Conclusions (I): PosiDve and negaDve aspects of passion

There are hardly any individuals with stronger OP than HP. Strong HP without strong OP happens, but more oten, both experiences are aligned.

Passion is not alternaDvely obsessive or harmonious for the majority of individuals, but characterized at the same Dme by obsessive and harmonious aspects.

The desirable and undesirable aspects of passion are two sides of the same coin.

The harmonious experiences are stronger than the obsessive for most individuals.

This is in line with other recent findings that pointed out the co-‐occurrence of posiDve and negaDve aspects of strong moDvaDon (e.g. profiles of engaged-‐exhausted individuals, see Tuominen-‐Soini & Salmela-‐Aro, 2014; Salmela-‐Aro et al., revise and resubmit).

Conclusions (II): misleading z-‐scores -  Use raw scores or alternaDve scale transformaDons,

such as POMS (Liqle, 2013) in the analysis of profiles and groups! Z-‐scores are oten misleading.

-  3 problems:

SomeDmes z-‐scores suggest that response A was higher than response B, even though it was the other way around when the person answered to the original response scale!

z-‐scores can also suggest that an answer was “high”, even though this answer negated the item statement on the original response scale (=was a raDng below the scale midpoint on a scale from 1 = don’t agree at all to 7 = totally agree).

Plovng profile differences using z-‐scores oten equals plovng a graph with a truncated Y-‐axis, because rarely the whole area of possible answers is shown in such z-‐score graphs.

References Liqle, T. D. (2013). Longitudinal Structural Equa7on Modeling (Methodology in the Social Sciences). The Guilford Press. Mageau, G. A., Vallerand, R. J., Charest, J., Salvy, S.-‐J., Lacaille, N., Bouffard, T., et al. (2009). On the development of harmonious and obsessive passion: The role of autonomy support, acDvity specializaDon, and idenDficaDon with the acDvity. Journal of Personality, 77(3), 601-‐646. doi:10.1111/j.1467-‐6494.2009.00559.x Moeller, J., Keiner, M., & Grassinger, R. (accepted). Two sides of the same coin: Are the dual ‘types‘ of passion types in the sense of disDnct subgroups? Journal for Person-‐Oriented Research. Moeller, J. (revise and resubmit). A word on standardizaDon in longitudinal studies: Don‘t. Manuscript resubmiVed for publica7on to Fron7ers of Psychology. Philippe, F., Vallerand, R.J., & Lavigne, G. (2009). Passion does make a difference in people’s lives: A look at well-‐being in passionate and non-‐passionate individuals. Applied Psychology: Health and Well-‐Being, 1, 3-‐22. doi:10.1111/j.1758-‐0854.2008.01003.x Salmela-‐Aro, K., Moeller, J., Schneider, B.; Spicer, J., & Lavonen, J. (revise and resubmit). IntegraDng the light and dark sides of student engagement with person-‐oriented and situaDon-‐specific approaches. Invited manuscript submiVed for publica7on to Learning and Instruc7on. Tuominen-‐Soini, H., & Salmela-‐Aro, K. (2014). Schoolwork engagement and burnout among Finnish high school students and young adults: Profiles, progressions and educaDonal outcomes. Developmental Psychology, 50(3), 649–662. doi:10.1037/a0033898 Vallerand, R. J., Blanchard, C. M., Mageau, G. A., Koestner, R., Ratelle, C., Léonard, M., et al. (2003). Les passions de l'âme: On obsessive and harmonious passion. Journal of Personality and Social Psychology, 85(4), 756-‐767. doi:10.1037/0022-‐3514.85.4.756 Vallerand, R.J., & Houlfort, N. (2003). Passion at work: Toward a new conceptualizaDon. In D. Skarlicki, S. Gilliland, & D. Steiner (Eds.), Social issues in management (Vol. 3, pp. 175–204). Greenwich, CT: InformaDon Age Publishing. Wang, C. K. J., Khoo, A., Liu, W. C., & Divaharan, S. (2008). Passion and Intrinsic MoDvaDon in Digital Gaming. Cyberpsychology and Behavior, 11(1), 39-‐45. doi:10.1089/cpb.2007.0004 Wang, C.C., & Yang, H.W. (2008). Passion and dependency in online shopping acDviDes. Cyberpsychology and Behavior 10 (2), 296–298. hqp://dx.doi.org/10.1089/cpb.2006.9954.

Vielen Dank für Ihre Aufmerksamkeit!

3. Structure Julia Moeller 31.01.2013

Thank you for your abenFon and

feedback!

Contact: [email protected]

Special thanks to: Ullrich Trautwein, Rainer Silbereisen,

Katariina Salmela-‐Aro, Denise Wächter, Dieter Lippold ,

Thomas Münzberg, Atademes Branco Pereira,

Montgomery Miranda, Marina Pickler Rorato, Andréa Diniz Silva, &

Tobias Antunes de Amorim.

Why plovng z-‐scores in graphs about profile differences is misleading

DescripDves for the assessed scales Study 1 Study 2 Study 3 Study 4

M SD α M SD α M SD α M SD α

HP 5.15 .849 .73 5.29 .993 .69 5.48 .978 .89 5.35 .845 .73

OP 3.70 1.266 .85 4.41 1.476 .87 4.34 1.540 .93 3.79 1.291 .85

Passion criteria 5.95 .722 .58 5.74 .905 .50 6.13 .876 .85 6.00 .991 .84

Passion (com.pass) 4.45 .734 .91 4.63 .745 .88 4.74 .715 .96 4.50 .662 .91

Commitment 4.24 .595 .74 4.22 .699 .83

BAS 3.30 .417 .86 3.10 .375 .76

PosiFve Affect 4.00 .526 .75 4.16 .647 .91

NegaFve Affect 1.47 .423 .75 2.35 .997 .92

Dependency 3.14 .761 .73

Education

Two sides of the same coin: Are the Dual Types of Passion Types in the Sense of Distinct Subgroups of Individuals?