Over the last few decades national and international studies about romantic relationships have gained strength, such relationships are considerably serious outside marriage or cohabitation. Adolescents and young relationships, which are prior to the consolidation of the couple and outside marriage or cohabitation -known as dating- (Connolly & McIsaac, 2011), tend to be different from those held by adults in areas such as level of commitment, duration, sexual intimacy and the way to solve conflicts (Furman & Wehner, 1997; Molidor & Tolman, 1998). Thus, the violent dynamic that might arise will have different characteristics (for example, there's no financial dependence, emotional blackmail or other abusive conducts in relation to children, or household co-responsibility, etc.). The Report of Youth in Spain (INJUVE, 2012) points out that only 23.8% of young people between 20 to 24 years of age live with their partners and it is also observed that the higher the educational level the higher the percentage of youngsters living at the parental home. All these features make the relationships and violent manifestations among young university couples quite different from the adult ones. Violence in dating relationships in young people are characterized for being moderate, bidirectional and reciprocal (Nocentini, Pastorelli, & Menesini, 2011; Ortega & Sánchez, 2010; Viejo, 2014).

Notwithstanding, there has been less research on psychological violence than on other types of maltreatment, like physical or sexual abuse. Perhaps, the lack of psychological violence centered research is due to the fact that it can be less objective and more difficult to evaluate than physical maltreatment and other types of violence (Calvete, Corral, & Estévez, 2005; Rodríguez-Carballeira et al., 2005).

It has been in the last few decades when research interest has emerged in this field regarding adolescent and young couples’ relationships. The majority of studies that include this or any other type of violence in dating relationships have considered it as a risk factor of violence in the adulthood or marital couples (Gormley & López, 2010; Moreno-Manso, Blázquez-Alonso, García-Baamonde, Guerrero-Barona, & Pozueco-Romero, 2014), establishing that psychological partner violence is a behaviour repeated along the following relationships (Lohman, Neppl, Senia, & Schofield, 2013).

Scientific literature has established that psychological violence is defined by attitudes, behaviours and styles of communication based on humiliation, control, disapproval, hostility, denigration, domination, intimidation, threat of direct violence and jealousy (Murphy & Hoover, 1999; O’Leary & Smith-Slep, 2003). O’Leary (1999) identified in his definition control and domination actions but also verbal aggression including denigration and recurring criticism towards the partner. Marshall (1999) introduced a new perspective in the study of psychological violence by differencing overt and subtle ways of abuse. Overt psychological violence is characterized by spreading behaviors of control and dominance easy to recognize because an aggressive and dominant style is used and it clearly affects resulting feelings, including: domination, indifference, monitoring and discredit. This type of abuse tends to occur in situations of conflict. Nonetheless, subtle psychological violence can appear in loving, joking and caring situations. Messages and actions to undermine, discount and isolate the partner are defined as subtle. These forms are independent from domination and produce an emotional damage that is difficult to recognize as abusive.

International and national research on psychological violence has shown higher rates of prevalence than other types of intimate violence (Liles et al., 2012; Zorrilla et al., 2010). These higher rates of psychological violence have been identified in dating relationships in which the implication is around 80%. Percentages regarding victimization range between 76-87% among boys and 78-82% among girls, and regarding aggression between 74-85% among boys and 83-90% among girls (Cortés-Ayala et al., 2014; Hines & Saudino, 2003; Straus, 2004; Straus, Hamby, Boney-McCoy, & Sugarman, 1996). In a recent study with university students about psychological abuse, different types of behaviors were observed to define this phenomenon, such as disparagement, hostility, indifference, intimidation, imposition of behavioral patterns, blaming and apparent kindness; the results pointed out that the indifference was the most common form of psychological violence in dating (Blázquez-Alonso, Moreno-Manso, & García-Baamonde, 2012). Different studies have found gender differences in psychological violence. On one hand, females perpetrate significantly more psychological aggression than males (Hines & Saudino, 2003). On the other hand, many studies and social opinion supported by media establish a wider presence of psychological abuse manifestations with the highest evidence of a greater rate of patterns among men (Moreno-Manso et al., 2014).

It is widely recognized that psychological violence is fairly stable and has a severe impact (Carney & Barner, 2012; Shortt et al., 2012). Marshall (1999) recognized that subtle and overt acts of psychological abuse are likely to harm self-perception, well-being and the perception of the relationship and the partner. Research has shown that psychological and physical violence are interrelated, actually psychological violence could precede physical one (Muñoz-Rivas, Graña, O’Leary, & González, 2007; Straus et al., 1996).

The most recognized instruments about psychological abuse in dating relationships are The Conflict in Adolescent Dating Relationship Inventory (CADRI) (Wolfe et al., 2001), validated with Spanish sample (Fernández-Fuertes, Fuertes, & Pulido, 2006); and the Modified Conflict Tactics Scale (M-CTS) (Neidig, 1986). Both scales consider whether the person had suffered or perpetrated different violent behaviors. M-CTS and CADRI included psychological violence as a short subsection into the dating violence questionnaire, dismissing the individual nature of the phenomenon. The M-CTS has been criticized because it does not include restrictive behaviors and public humiliation (Calvete et al., 2005).

Due to these limitations some studies have decided to use measures which had been designed for adult population. There is a wide range of instruments designed to measure psychological violence from a women maltreatment perspective in adulthood, as the Inventory of Psychological Abuse in the Context of Couple Relationships (Calvete et al., 2005), The Index of Spouse Abuse (Hudson & McIntosh, 1981) and The Spanish Version of the Index of Spouse Abuse (Sierra, Monge, Santos-Iglesias, Bermúdez, & Salinas, 2011). The Psychological Maltreatment of Women Inventory (PMWI) (Tolman, 2001) is one of the most used scales to evaluate abuse against women; it has two subscales called domination-isolation and emotional-verbal aggression. The Scale of Emotional Abuse (Murphy & Hoover, 1999) offers four subscales: Domination-intimidation, Restrictive isolation, Denigration and Hostile withdrawal and The Profile of Psychological Abuse (Sacket & Saunders, 1999) is composed of four victimization factors referred to ridicule traits, criticize behaviours, to ignore the partner and control-jealousy. Marshall created the Subtle and Overt Psychological Abuse of Woman Scale (SOPAS). The psychometrics properties of this scale were presented by Jones, Davidson, Bogat, Levendosky and Von Eye (2005); and Buesa and Calvete (2011) adapted this scale to Spanish population finding that violence against women by an intimate partner can take many modalities, including forms of overt and subtle victimization in only one factor.

Nevertheless, most of these instruments have a gender bias -male violence against women-, considering only women or women victims of domestic violence perspectives. Therefore its generalization is questioned in social areas in which some domesticity elements are not present, such as house sharing, children responsibilities or shared properties. Besides, most of these instruments focus on the victimization of the questioned person, leaving aside the possibility that the victim might also be an aggressor, thus excluding an important factor: the bidirectional or reciprocal violence dynamic, which has been identified among adolescents and young people in numerous national and international studies. These studies have pointed out the reciprocal relationship between victimization and aggression in psychological dating violence (Fernández-González, O’Leary, & Muñoz-Rivas, 2013; Menesini, Nocentini, Ortega-Rivera, Sánchez, & Ortega, 2011; Orpinas, Nahapetyan, Song, McNicholas, & Reeves, 2012) and more specific instruments of psychological violence are needed to study the bidirectionality in dating couples and to evaluate different types of behaviours and attitudes, both subtle and overt.

The objectives of this study are: a) to develop the scale Psychological Dating Violence Questionnaire (PDV-Q) with a sample of university students and b) to analyze its reliability and validity. It is hypothesized that will be tested a bi-dimensional structure with aggression and victimization scales.

ResultsExploratory Factor Analysis

An EFA was done to explore the number of scales and it was observed a clear differentiation between two factors: Aggression and Victimization, according with the theory. Later on, exploratory tests were made with a different number of factors but results were not conclusive. Keiser-Meyer-Olkin test and Barlett test showed satisfactory results for the double factor solution and revealed an optimum adequacy of the correlation matrix (KMO=.78), Barlett test result was p < .01 (Delgado-Rico et al., 2012). The total observed variance found was 38%. The majority of the items acquired values above .40 (Neukrug & Fawcett, 2014). Taking into account communalities and factorial weight values, according to the established criteria, items 1, 3, 11 and 19 for victimization and aggression were removed. The values found in the EFA are shown in the Table 1.

Table 1.

Univariate descriptive for the EFA.

Items	M	SD	Skewness	Kurtosis	Loadings	h2	R
Victimization
Item 1	1.53	1.73	0.24	-0.50	.34	.85	.49
Item 2	0.79	0.97	1.17	0.64	.38	1.00	.58
Item 3	0.35	0.49	2.64	7.15	.63	.90	.71
Item 4	0.51	0.67	1.79	2.83	.60	1.00	.63
Item 5	1.30	1.49	0.28	-0.79	.40	.91	.59
Item 6	0.48	0.63	1.87	3.70	.78	.92	.68
Item 7	0.79	0.97	1.14	0.46	.43	1.00	.68
Item 8	0.25	0.37	3.20	11.86	.81	1.00	.70
Item 9	0.34	0.50	2.58	5.59	.54	1.00	.65
Item 10	0.57	0.77	1.81	2.22	.64	1.00	.63
Item 11	1.66	1.89	0.14	-1.13	.38	.99	.42
Item 12	0.71	0.89	1.25	0.66	.58	1.00	.74
Item 13	0.75	0.94	1.26	0.61	.65	.89	.63
Item 14	0.83	1.03	1.21	0.61	.63	.92	.66
Item 15	0.92	1.13	1.18	0.53	.38	1.00	.59
Item 16	0.66	0.49	1.53	1.58	.65	1.00	.61
Item 17	0.66	0.48	1.53	1.57	.64	.89	.57
Item 18	0.34	0.20	2.28	5.44	.67	.95	.78
Item 19	1.35	1.16	0.44	-0.63	.37	.70	.50
Agression
Item 1	1.63	1.83	-0.04	-0.62	.39	.78	.33
Item 2	0.71	0.88	1.12	0.50	.46	.94	.50
Item 3	0.25	0.36	2.94	9.72	.39	1.00	.64
Item 4	0.42	0.55	1.82	3.01	.41	1.00	.54
Item 5	1.25	1.43	0.54	0.08	.48	.99	.52
Item 6	0.47	0.62	1.71	2.76	.68	.86	.67
Item 7	0.70	0.86	1.22	1.31	.60	.91	.60
Item 8	0.19	0.29	3.03	9.75	.74	1.00	.73
Item 9	0.16	0.26	3.58	13.07	.37	1.00	.58
Item 10	0.47	0.66	2.37	4.83	.56	1.00	.59
Item 11	1.64	1.86	0.16	-0.99	.42	.80	.30
Item 12	0.72	0.89	1.13	0.19	.37	.92	.58
Item 13	0.72	0.89	1.11	0.33	.42	.92	.55
Item 14	0.65	0.81	1.21	0.55	.45	.84	.61
Item 15	0.86	1.06	0.99	-0.07	.46	.91	.47
Item 16	0.72	0.90	1.24	0.77	.77	.97	.55
Item 17	0.74	0.93	1.42	1.34	.75	.97	.47
Item 18	0.29	0.41	2.16	4.18	.40	.97	.68
Item 19	1.43	1.64	0.35	-0.80	.34	.82	.38

Confirmatory Factor Analysis

A Confirmatory Factor Analysis was carried out. Items 8 and 17, with correlation values higher than .8 were removed. Finally the questionnaire consists of 13 items for victimization and 13 for aggression (see Appendix). The CFA results confirmed the bi-dimensional model from the EFA (see Figure 1 and Table 2).

Figure 1.

Standardized factor loadings of Confirmatory Factor Analysis.

(0.31MB).

Table 2.

Model fit indices Confirmatory Factor Analysis.

p	χ2/df	CFI	GFI	TLI	RMSEA	RMR	ECVI
.000	2.03	.99	.99	.99	.06	.05	.08

Reliability analysis

The reliability analysis showed Cronbach's values of .88 regarding victimization subscale, .85 regarding aggression subscale and a total value of .92 (Neukrug & Fawcett, 2014).

External validity

Correlation analyses between psychological and physical forms of violence were run to measure the external validity of the PDV-Q. The results showed that psychological aggression was positive and significantly related with physical aggression (.51) and physical victimization (.48). Psychological victimization was also significantly related with physical aggression (.42) and physical victimization (.48). Nonetheless, all the correlation values were medium-low, pointed out the differences between both concepts.