Vivo Questionnaire: A Measure of Human Worldviews and Identity in Trauma, Crisis, and Loss—Validation and Preliminary Findings

Various proposals have been made in order to measure worldviews and identity in extreme trauma or loss experiences. The use of these scales has provided mixed results. The Vital Impact Assessment Scale (VIVO) was designed to fill some of the gaps of the existing tools: (a) to be administered to the general population, as well as to survivors of extreme experiences; (b) to include an extensive range of human responses to extreme situations; (c) to be administered to large population samples from different cultural backgrounds; (d) to avoid anchoring to a specific personal experience; and (e) to work with extensive and complex response profiles. Validation data and preliminary results are presented for the Spanish version, and an English formulation is proposed. The final version of the VIVO Questionnaire, composed of 116 items, offers an organized profile divided into 10 conceptual blocks (worldviews, attitude towards the world, view of human beings, coping, impact of past situations, emotions, telling the experience, consequences, social support, and identity) and 35 subscales.

Validation data and preliminary results are presented for the Spanish version, and an English formulation is proposed. The final version of the VIVO Questionnaire, composed of 116 items, offers an organized profile divided into 10 conceptual blocks (worldviews, attitude towards the world, view of human beings, coping, impact of past situations, emotions, telling the experience, consequences, social support, and identity) and 35 subscales.
Various proposals have been made in order to measure worldviews in extreme trauma or loss experiences. The most frequently used measures have been the World Assumption Scale (Janoff-Bulman, 1992), the Post-Traumatic Cognitions Inventory (Foa, Ehlers, Clark, Tolin, & Orsillo, 1999), and the Changes in Outlook Questionnaire (Linley & Joseph, 2004).
The use of these scales throughout the past 10 years has provided mixed results. They have revealed significant clinical data obtained from general samples (Startup, Makgekgenene, & Webster, 2007), and especially in survivors of interpersonal violence (Ali, Dunmore, Clark, & Ehlers, 2002), sexual abuse (Harris & Valentiner, 2002;Ullman, 1997), the Holocaust (Magwaza, 1999), or political violence. But many other studies have failed to find differences between clinical and nonclinical populations, or data from worldviews have correlated poorly with PTSD and other trauma response measures (Jeavons & Godber, 2005;Kaler et al., 2008).
Tools of this type face some particular epistemological difficulties: (a) They try to measure constructs that are dynamic, elusive, and dialectical with the environment; (b) the person often lacks clear insight and a global vision of his or her worldviews; and (c) while the purpose of the items is to capture abstract or existential concepts, the person usually answers by resorting to personal or vicarious situations that are considered to be prototypical (Tourangeau, Rips, & Rasinski, 2000). When these situations do not exist or cannot be recalled at the moment, people respond from a purely speculative point of view or based on an estimate of how they consider that they would react, which corresponds to a desideratum (''wishful thinking'') or to a socially desirable answer (Sumalla, Ochoa, & Blanco, 2008). Worldview measures have not been designed to be applied to the general population, and studies must compare groups of people with experiences of different levels of intensity (Solomon, Iancu, & Tyano, 1997). This goes against accumulated knowledge indicating a profound individuality in trauma responses and a lack of a dose-response connection between the type or intensity of the traumatic event and changes in worldviews (Basoglu & Parker, 1995;Fujita & Nishida, 2008). It is also important to add the difficulties of establishing causal inferences from studies that are cross-sectional (Kaler et al., 2008) and do not provide follow-up data to determine the sensitivity of worldview measures to change.
The Vital Impact Assessment Scale (Escala de Valoración del Impacto Vital, or VIVO) has been developed with various purposes: (a) to provide a scale that can be administered to the general population, as well as to those who have survived extreme experiences, in order to obtain comparative data; (b) to take into account an extensive range of human responses to extreme situations; (c) to provide an instrument that can be administered to large population groups from different cultural backgrounds; (d) to avoid anchoring to a personal experience; and (e) to work with ''response profiles'' instead of with added scores. Validation of the VIVO scale has been carried out according to the standards for the development and review of instrumental studies, and the results are presented herein.

Samples and Recruitment Procedures
The questionnaire has been developed during the past 4 years. Initial drafts of the VIVO Questionnaire (2005)(2006)(2007) were tested with ad hoc samples of the general population, psychologists, firemen, and ambulance drivers, as well as inpatients in burn units at a general hospital and outpatients in the department of psychiatry in a complex trauma unit. The scale was implemented in a paper version, and individual interviews were conducted to check item by item for content validity and suggestions. The VIVO was also included on a Web page in May 2007 asking for voluntary participation. Participants (n ¼ 189) were offered, through a form, the opportunity to explain their answers to every item and to suggest additional wording and potential unexplored issues according to their life experiences. Both paper and Web questionnaires were compared item by item and via global scores (t tests), yielding no significant differences after adjusting for multiple comparisons.
The final scale sample was recruited through the project's Web page (http://www.psicosocial.info/) from September 2007 to February 2008 (n ¼ 3130; Version 1). Preliminary analysis showed the convenience of changing the anchoring system of the items and launching a second wave (March 2008-October 2008n ¼ 827;Version 2 [final]). It was decided to obtain samples using the Internet as a strategy to obtain a number of responses large enough to have some key traumatic events well represented and to have big samples from different countries and different cultures. Therefore, voluntary collaboration was requested in approximately 1,400 Spanish-language Internet forums. The forums were randomly selected using search engines, and they correspond to the endless and heterogeneous scope of subject matters represented in these types of public spaces (a complete list is available on request). The questionnaires obtained were later analyzed one by one by three independent judges, ruling out acquiescent answers or answers with incoherent or extreme profiles (outliers) (n ¼ 314; 4.9%).
The final refined sample was composed of 3,808 people (3,020 for Version 1 and 788 for the slightly modified Version 2). From a demographic point of view, there were no differences between the questionnaires that were ruled out and those that were finally analyzed. Despite the self-selection bias that this type of procedure may exhibit, its correct use together with individualized screening filters has been widely accepted as an adequate and valid contemporaneous sampling method (Eiroá-Orosa, Fernández-Pinto, & Pérez-Sales, 2008;Fortson, Scotti, del Ben, & Chen, 2006). Our online questionnaire followed the technical recommendations provided by the main Web-based studies available to date (Dillman, 2007).
The final sample was obtained from 12 countries, especially from Spain (32.2%), Argentina (19.8%), Mexico (11.4%), Chile (10.1%), and Colombia (9.6%). The mean age was 28 years (SD ¼ 10.79), and 51% were women. The majority were single (63.1%), had a high level of education (62%), and described themselves as of a middle-class socioeconomic level (63.1%). Additionally, 34.5% considered themselves to be of left-wing ideology, 25.5% positioned themselves in the center, and 20.9% were of right-wing ideology; 57.5% did not practice any religion, and of those who described themselves as religious, most were Catholic (32%).
Drafts of the VIVO Questionnaire have been used in studies with relatives of people who were detained or disappeared for political reasons in Argentina (Arnoso-Martínez & Eiroa-Orosa, 2010), victims of massacres in Colombia making claims for reparation, and victims of torture, asylum seekers, and adolescents with antecedents of child sexual abuse in Spain (unpublished results).
The project followed the standards for carrying out investigations with questionnaires that have been drawn up by the Council of American Survey Research Organizations (2004) and the regulations set by the Spanish Agency of Data Protection, and it was approved by the Ethics Committee for Basic and Clinical Research of La Paz University Hospital (Madrid).

Measures
With respect to the VIVO Questionnaire, an initial pool of 128 expressions in Spanish was derived from survivors' testimonies and expert consensus corresponding to basic beliefs; signs of subjective damage, resilience, and growth; social, cognitive, and behavioral appraisal and processing of extreme experiences (Pérez-Sales, 2006); and an extensive review of existing models and measures focusing on the impact of extreme experiences (Antonovsky, 1987;Calhoun & Tedeschi, 1999;Ehlers & Clark, 2000;Epstein, 1989;Janoff-Bulman, 1992;Kauffman, 2002;Kobasa, 1983;Pennebaker, 1990;Stroebe, Schut, & Finkenauer, 2000), ethnographic and qualitative studies (Lifton, 1967;Pérez-Sales, Bacic, & Durán, 1998) or autobiographical accounts of victims' experiences (Amery, 2001;Frankl, Lasch, & Allport, 1963;Levi, 1987;Steinberg, 2001). A total of 80 items were suitable for the general population, and 48 were meaningful only for survivors. In earlier versions, the participants had to choose between opposing expressions of a concept. This was later changed to a classic 5-point Likert scale. Although we present results for Version 2 (n ¼ 788) and the final sample, all statistical analyses were repeated independently of the previous one (n ¼ 3,020), and the results produced were almost identical.
In addition to the VIVO Questionnaire, participants completed the following measures.
. Inventory of Extreme Experiences (IEE) (Pérez-Sales, Cervellón, Vázquez, Vidales, & Gaborit, 2005): This instrument collects data on 24 experiences (most commonly linked to trauma, loss, or crisis and five positive life events). . PTSD Checklist-Civilian Version (PCL-C): a 17-item scale that assesses posttraumatic stress disorder (Weathers, Huska, & Keane, 1991). Different cutoff points have been proposed (Vazquez, Pérez-Sales, & Matt, 2006), but only the overall score was used in this study. . Overall satisfaction with life: ranked from 1 (I consider myself to be happy with life) to 3 (I consider myself to be unhappy overall). . Overall self-image: ranked from 1 (I like the way I am and I would change very little) to 3 (There are more things that I don't like about myself than things that I do). . Average positive, negative, and overall mood: Emotions of sadness and joy=happiness were evaluated (1 ¼ constantly, 5 ¼ never) for the previous 2 weeks, as well as overall mood (1 ¼ euphoric and full of vitality during most of the day, 5 ¼ very sad during most of the day).

Test for Normality and Item Distribution
Descriptive statistics and the actual distribution of responses were computed for each VIVO item. Means ranged from 1.84 to 4.33 (full data available on request). Most of them fell around the expected value (mean ¼ 3), and all were in the range of one standard deviation from the theoretical mean (1.4-4.6), had a skewness of 2 or below, and had a kurtosis of 7 or below (Fabrigar, Wegener, MacCallum, & Strahan, 1999).

Factor Structure
The items were split into 10 conceptual blocks according to the results obtained from preliminary studies (Pérez-Sales et al., 2005), the theoretical framework of the model, and an expert consensus. Five blocks correspond to the 80 items suitable for the general population and survivors, and five blocks correspond to the 48 items only suitable for survivors. We performed factor analyses using principal component methods with varimax rotation in each of these 10 conceptual blocks. Thirty-five factors were obtained (see Table 1) that explained between 44.1% (coping) and 56% (emotions associated with what happened) of the variance. There were seven items with low factor loadings in 4 of the 35 factors (''tolerance of ambiguity and uncertainty,'' ''search for logic,'' ''capability of communicating what happened,'' and ''development of victim identity''); these were retained because previous results showed their conceptual contribution to the scale's global interpretation.
The structure of these 35 factors represents Cronbach alpha values from 0.6 to 0.83, with the exception of the four previously mentioned factors, which have lower values that range from 0.33 to 0.38.
Five items were reworded because they showed a low factor structure fit or favored double denial expressing ambiguity. In addition, due to the fact that some factors were only composed of a few items, the appropriateness of presenting the items in mixed order was evaluated in order to increase reliability and avoid bias and entrainment effects. Table 2 provides the definition of the factors included in the VIVO scale. An analysis using Spearman's correlation of the 35 factors showed values ranging between 0.001 and 0.583. Viewed overall, 441 (36%) showed null or very low correlations (between 0 and AEÀ0.2), 686 (56%) showed low correlations (between AEÀ0.2 and AEÀ0.4), and 98 (7.7%) showed moderate correlations (between AEÀ0.4 and AEÀ0.58).

TEST-RETEST
In May 2008, all participants who had completed the VIVO in the period October-December 2007 were contacted. A total of 186 participants agreed to complete the questionnaire again. Paired item-by-item t tests showed stability in 125 of the 128 items. Table 3 shows the correlations between the different subscales and the external criteria on validation. In 23 of 25 subscales, there are significant, but moderate, correlations between vital impact subscales and posttraumatic stress measures. Similar patterns with moderate correlations can be observed between vital impact measured with the VIVO scale and self-image (28=35), satisfaction with life (28=35), sadness (31=35), happiness (27=35), and overall mood (29=35). In the VIVO subscales, lower scores are associated with negative impact, damage, or management difficulties, which is why the negative correlations with the criteria variables indicate that people who have a better Note. Values are means, standard deviations, and internal consistency for each of the factors for the general population (1-21) and survivors (22-35). VE ¼ variability explained by the factor structure of each conceptual block; a ¼ safety coefficient (Cronbach a of each factor). a Items retained based on theoretical criteria.

242
P. Pe´rez-Sales et al.  . . . tend to look for logic=rational in life situations and ask themselves why things happened.
Conceptual Block III -View of Human Beings Sharing the experience (Useful=Not useful) . . . consider that talking about the experience is not very useful and that silence or leaving things behind are better strategies.
. . . consider that sharing the experiences helps and that it provides a sense of relief.
Significance of suffering (Useful= Not useful) . . . consider that suffering is useless and that it ruins people.
. . . consider that suffering is an opportunity for learning and overcoming situations, and that it is possible to be happy even while suffering. Human kindness (Present=Absent) . . . consider that evil generally triumphs and that people don't tend to help those who are close to them.
. . . consider that there is a human tendency towards kindness.
Finding words (Possible=Not possible) . . . consider that there are no words to express the horror and that if there were, it still wouldn't be understood by others.
. . . consider that there are always words to express even the most horrific experiences.

Dreams (Neutral= Suffering)
. . . consider that their suffering is expressed in their dreams.
. . . usually do not remember their dreams and tend to not give importance to them.
244 P. Pe´rez-Sales et al. . . . consider that they have painful feelings of guilt.
. . . consider that they can assume responsibility for the past or that it is possible to understand or forgive others. Self-confidence (Intact=Lost) . . . consider that they have lost confidence in themselves and their capability of confronting their problems.
Learning (Possible= Not possible) . . . consider that it is not possible to learn from one's mistakes.
. . . consider that they have learned or that they have become stronger from the adverse experiences. Belief in the possibility of change (Possible= Not possible) . . . consider that it is not possible for human beings to change.
. . . .consider that changes are part of being human.

Fears (Specific= Unspecific)
. image of themselves, who are more satisfied with the way their life is going, and who show more happiness and a better overall mood and state of mind also tend to have higher VIVO subscale scores. The positive correlation with sadness indicates, as expected, that feelings of sadness are associated with lower scores on most of the VIVO subscales. The scales that do not show a negative correlation with posttraumatic stress, satisfaction, and overall mood measures are, constantly and homogeneously (see Table 3), the ones related to attitudes towards the world (political or religious convictions, belief in destiny, ambiguity and uncertainty, search for logic in life situations) and to ''finding the words'' (sharing the experience, leaving behind as a way of coping, value of testimony, communicating what happened, sensitivity-insensitivity towards others). This indicates that both types of variables are independent regarding the vision that the person has of others, of him-or herself, and of the world. Additionally, in relation to self-image, there are some expected negative correlations, particularly under the specific subscales related to survivors, experiencing identity changes, and changes in priorities (Table 3).

DISCUSSION
We have presented a scale that analyzes worldviews; assesses the impact caused by vital experiences of trauma, loss, or crisis; and can be administered to the general population and to survivors. The scale was developed based on a series of studies that have been carried out by our team throughout the past 4 years. During its first versions (2006)(2007), we carried out validation processes on clinical and nonclinical samples and test-retest analyses after the first 6 months, showing good results. We compared paper-and-pencil and Web samples showing the adequacy of both methods for the purpose of validating the VIVO. Using a large sample (n ¼ 788) backed up by another sample obtained with an identical older version, except for type of anchoring (n ¼ 3,130), we obtained a structure with 10 conceptual blocks and 35 subscales that covered areas related to worldviews, attitude towards the world, view of human beings, coping, impact of past situations, emotions, communication, consequences, social support, and identity. The result was a final questionnaire of 116 items (see the appendix). The general population answers only the first 72 items, and survivors go on to answer the remaining 44. The estimated time to complete the questionnaire ranges from 30 to 45 minutes, depending on the person's level of education. The questionnaire is complex, not so much because of the questions or the phrasing (in previous versions and based on different patients and groups of people, the majority of complex terms or ambiguous expressions have been eliminated), but because of the profoundness of the aspects explored, which require a certain introspective effort. As a result, although the questionnaire has shown excellent clinical results and it has proven to be very useful in terms of psychotherapy-related work with survivors in general, it presents some difficulties when working with indigenous or elderly people.
Psychometrically speaking, we have chosen a solution with a high number of short subscales (35) consisting of 2 to 12 items per scale. We found good values of explained variance and reliability coefficients within acceptable ranges, apart from some exceptions that were retained based on theoretical criteria. Several studies have shown that scales with few items, such as the ones that make up the VIVO Questionnaire, can be solid and useful measures (Cappelleri et al., 2009). There is a growing tendency in contemporary research to favor using multiple scales with a reduced number of items instead of a short array of very long measures. The worldviews of a person cannot be reflected by a global score of partial aspects. The positive or negative impact on a sole nuclear aspect out of the many aspects explored by VIVO can change a person's life, and this is critical to our proposal of working with extensive profiles addressing the key components of 248 P. Pe´rez-Sales et al.
a survivor's experience. In conclusion, throughout the past few years we have attempted to find a delicate balance between an instrument that explores the complexity of existential responses to traumatic situations of loss and crisis and one that can also be simple and manageable enough to be used in research and in everyday clinical activities.
The result is not a group of values (it is complicated to interpret each one of them separately), but a profile. From a clinical point of view, the VIVO Questionnaire helps to easily detect key issues and areas to be explored in psychotherapy (Pérez-Sales, 2006). Figure 1 shows the profiles for two persons, the first one with no significant traumatic background and the second corresponding to a woman from Colombia whose husband was murdered.
The correlations between the VIVO scales and the PCL-C are moderately significant (the values are not higher than 0.5), which supports the idea that although the impact of extreme experiences on worldviews is related to PTSD, both constructs are sufficiently different and add complementary views to our understanding of human responses to trauma. While PTSD, generally speaking, can be conceptualized as a physiological response related to the biological circuits of fear and to the difficulties that arise from the fact that the memories associated with the disorder are registered into a coherent sequence of events in one's episodic memory (Shin & Handwerger, 2009), measures on vital impact deal with the most profound aspects of a human being's experiences, which are often linked to nonconscious information processing. As long as the answers to questionnaires are based on reflexive processes, they will have to be complemented with research using experimental measures of unconscious information processes (selective attention neuropsychological tests, response systems under time pressure, etc.). Finally, scores on the VIVO scales linked to resilience show a high direct correlation with measures of self-image, satisfaction with life, and a positive mood but not with sadness, in line with previous studies (Tomich & Helgeson, 2002;Wilson, 2006), with the exception of subscales related to communication of experiences and attitudes towards the world, as also pointed out in a validation study of the World Assumptions Scale (Kaler et al., 2008).
Due to the cross-sectional nature of our data, it is not possible to state whether being in a better mood led to answering the questionnaire with more benign responses or vice versa. Until this is clearly determined, our data suggest that when using the VIVO Questionnaire, it might be beneficial to include an emotional state measure and to control its possible influence as a covariable.
In short, the VIVO Questionnaire adds some innovative new features and concepts that can help to explore hypothetical relations between certain types of experiences and certain impact profiles in survivors. The subscales have shown high internal consistency, a well-justified structure derived both from theory and results, and adequate concurrent and discriminant validity. Our team is now conducting research on the impact of interpersonal violence versus other types of violence, the differential impact of trauma and loss on worldviews throughout the lifespan, and the specific effect on worldviews of political violence, just to provide some examples. We believe that relevant fields can be explored for comprehending the experiences that human beings go through in extreme situations.
Pablo Olivos works as a social psychologist at the University of Castilla-La Mancha and is a member of the Community Action Group. His research interests cover applied social psychology and social identity subjects such as environmental identity and restorative effects of contact with nature.
Elena Barbero-Val is a psychologist and member of the Community Action Group. She is currently preparing for the entry exam to the postgraduate training program in clinical psychology.
Alberto Fernández-Liria is a psychiatrist, chief of the University Hospital Principe de Asturias Mental Health Network, and associate professor at the Alcala University in Madrid, Spain. He is a member of the Community Action Group.
María Vergara works as a child psychologist at the EXIL Center, a program for medical, psychological, and social support for victims of violence and human rights violations.
La vida tiene sentido y por eso hay que vivirla. Life makes sense and that's why it has to be lived.

35.
Me paso el día dándole vueltas a algunas cosas que me han pasado. I spend a lot of time thinking about things that have happened to me.

36.
En realidad nunca se aprende de los errores. You never really learn from your mistakes.

37.
A veces tengo miedos que no logro identificar. Sometimes I have fears that I can't seem to identify.

38.
No me suelo acordar de lo que he soñado o si me acuerdo no le doy importancia. I don't usually remember my dreams and if I remember them I don't usually find them important.
There are always words to describe even the most horrible things.

40.
Aunque haya sufrimiento puede haber momentos de felicidad. Even when there is suffering there can be moments of happiness.

41.
He aprendido de mis errores en esta vida. I've learned from my mistakes in life.
Forgetting depends on yourself.

43.
Siento que las culpas no me dejarán de doler mientras no haya un castigo. I feel that in order to get rid of my guilt, I need to be punished.

44.
Olvidar no se elige. You don't choose to forget.

45.
El destino determina nuestras vidas. Our lives are predetermined by destiny.

46.
Mis convicciones ideológicas me han sido de ayuda en momentos difíciles. My ideological convictions have been of great help to me in difficult times.

48.
Es imposible sentirse seguro en esta vida. It's impossible to feel safe and secure in this life.

49.
Todo sufrimiento es una oportunidad de superación. All suffering is an opportunity for growth.

50.
No siento que pasar por experiencias horribles me hagan más fuerte como dicen. I don't think that living through horrible experiences makes me stronger, like people say.

51.
Siento que hay experiencias duras de la vida que me han hecho ser mucho más fuerte. I believe that I've been through some tough experiences, which have made me stronger.

52.
La gente no suele ayudar a los que están cerca. People don't usually help those close to them.

53.
Cuando me ocurre algo grave que me afecta mucho tiendo a bloquearme en ese momento. When something serious happens to me, I tend to freeze up.
Talking about things relieves suffering.