Psychometric Properties of the Dyadic Coping Inventory: Systematic Review

Dyadic Coping (DC) is a construct derived from Individual Coping that is based on one of the approaches of the systemic perspective, the Transactional Model. This model understands stress as a result of insufficient personal resources to deal with an aversive context. To deal with the negative effects of stress, cognitive styles and/or strategies are developed to face the event, aiming at emotional regulation to overcome the problem. Coping is understood as the efforts employed to manage stress (Lazarus & Folkman, 1984).

According to Story and Bradbury (2004), the analysis of Individual Coping may not be enough to deal with broader contexts, such as the marital one. Carr et al. (2014) highlighted that even when stress initially affects one member of the couple, its effects could have repercussions on both. Therefore, the emotional state of one would influence the emotional state of the other and bring implications to the stress level of the couple (Buchanan et al., 2012).

Bodenmann (1997) expanded Lazarus and Folkman (1984) individual Systemic Transition Model (STM) to the context of couples. Briefly, the STM for couples requires a few evaluative and coping steps: 1) one or both partners interpret a potential stressor and compare them with their resources and capacities (personal or environment); 2) if the stressor is understood as bigger than the resources that person has, the stress experience occurs and both verbal and non-verbal responses are issued to the partner; 3) from this communication and the interpretation of the stress, coping strategies are created for the stressed partner. In case the stress is perceived simultaneously by both partners either the dyadic strategy for coping or mutual support occurs.

To deal with stressful situations, the couple's members would develop interactional forms of coping, the Dyadic Coping. DC can either be positive or negative in nature (Bodenmann, 2005). According to the Systemic Transactional Model, the positive responses are: The Supportive Dyadic Coping (SDC), which occurs when one of the partners seeks to help the other to deal with some difficult situation, for example, by providing support and advice; the Common Dyadic Coping (CDC) which refers to a proportionality of engagement in actions among the members of the dyad to overcome an adversity; Delegated Dyadic Coping (DDC) which is the way in which one of the partners, upon the request for help of the other, takes on tasks/responsibilities to reduce the stress of their partner (Bodenmann et al., 2011).

Negative Dyadic Coping (NDC) strategies include the Hostile Negative Dyadic Coping, understood as aid accompanied by hostility and/or belittling of the partner; Ambivalent Negative Dyadic Coping, expressed by unwillingly help to the partner and/or from the perspective of personal gain; and Superficial Negative Dyadic Coping, when the support provided by the partner is not sincere or empathetic (Bodenmann & Randall, 2012). For Pires (2011), when Coping occurs through positive strategies, there are higher rates of couple’s well-being and health, with features such as stability and longevity of the relationship. NDC strategies, on the other hand, favor the emergence of new conflicts or aggravation of existing ones and shorten the relationship.

From the Systematic-Transactional perspective, the Fragebogen zur Erfassung des Dyadischen Copings als Tendenz (FDCT-N; Dyadic Coping Questionnaire) was developed to assess the marital patterns of Coping (Bodenmann, 1997, 2000). It consists of 68 items with a five-point scale, which measure the quality of communication, their own behavior and that of the partner in view of each other's perception of stress and the joint coping with the stressful situation. FDCT-N had two items related to the assessment of a person’s Coping and that of the partner (Bodenmann et al., 2018).

After FDCT-N, Bodenmann (2008) developed a reduced version of this instrument, the Dyadic Coping Inventory (DCI). It is a self-report instrument that assesses the quality of communication and Dyadic Coping behaviors that occur in an intimate relationship when one or both partners are stressed. It is composed of three scales that refer to the respondent's perspectives, Oneself (self-perception), Partner (partner's perception) and Common (couple's perception). The Oneself and Partner items and factors are similar, but modified to suit these different perspectives. In these scales, the following factors are found: Stress Communication, how I evaluate my own communication of stress (Oneself) and that of my partner (Partner); SDC, how I support my partner (Oneself) and how my partner supports me (Partner); DDC, how I incorporate my partner's tasks and assignments to reduce their stress (Oneself) and how my partner does it for me (partner); NDC, how hostile, insincere and willing I am to my partner when I go to help him (Oneself) and how my partner acts that way towards me (Partner). Finally, the Common Scale refers to the respondent's perception of the way he and his partner deal with stress together.

Since then, 15 studies of the DCI psychometric properties were carried out in several countries (Ekimchik & Kryukova, 2017; Falconier et al., 2013; Fallahchai et al., 2019; Gmelch et al., 2008; Kanth et al., 2022; Kurt & Akbaş, 2019; Ledermann et al., 2010; Levesque et al., 2014; Martos et al., 2012; Randall et al., 2016; Rusu et al., 2016; Shujja et al., 2020; Vedes et al., 2013; Wendołowska et al., 2020; Xu et al., 2016).

For the Standards for Educational and Psychological Testing, validity has a unique concept, which refers to the degree to which evidence support the interpretation of test scores. Accumulation of validity evidence reveals the justifiable condition for the use of a given score. There are four sources of validity evidence: content, internal structure, response process, relation to variables from other constructs and consequences of testing (American Educational Research Association [AERA], American Psychological Association [APA], & National Council on Measurement in Education [NCME], 2014).

This systematic review searched the literature for studies published on psychometric properties of the Dyadic Coping Inventory. This research study aims to contribute to clinical practice and research in marital and family relationships. It is an instrument empirically capable of detecting changes in the Dyadic Coping patterns, allowing clinicians to track the progress of the psychotherapeutic interventions using the DCI as one of the tracking indicators (Gmelch et al., 2008). For relationship researchers it provides evidences of the construct across different cultures and genders (Wendołowska et al., 2020).

Method

Systematic review is a scientific investigation that uses explicit and systematic methods with eligibility criteria to answer a given research question. For this purpose, it is necessary to have search methods, study selection criteria, methods for data extraction, analysis and synthesis, results, discussions, and final considerations. It is recommended that the review should have a protocol and be performed by a team working independently (Higgins et al., 2021). The Preferred Reporting Items for Systematic Reviews and Meta-Analyses checklist (Page et al., 2021) was used to guide the writing of this report.

Results

A total of 722 records were found in Capes’ Portal Periódico and on the Google Scholar Platform. After excluding duplicate materials on each platform and overlapping between the platforms, and after the application of the inclusion criteria, 10 studies from the portal and four from Google Scholar were included. After the last search update, another eligible study was found on Google Scholar, totaling 15 studies (Figure 1). Full access to this latest study was via Portal.

Portal provides a list of versions (repeated material) of publication of studies in each database. The accepted publications were counted only once, and the databases considered in order of results. After the exclusion of the repetitions, of the 15 databases, two were considered: Web of Science and Scopus, with 10 studies. In Google Scholar, four more databases were included, Directory of Open Access Journals (DOAJ), with two studies, DergiPark, Pubmed and Taylor and Francis Online, with three.

The studies are quantitative, cross-sectional, and empirical, except for Gmelch et al. (2008), and reviewed by peers, except for a journal (Psychiatry Hungarica) in which the information is not displayed. The publications were written in English except for Ekimchik and Kryukova (2017), Gmelch et al. (2008), Kurt and Akbaş (2019), and Martos et al. (2012). Only Falconier et al. (2013) informed the year of the research study: 2009. Five studies reported heterosexual orientation (Falconier et al., 2013; Gmelch et al., 2008; Levesque et al., 2014; Wendołowska et al., 2020; Xu et al., 2016).

In relation to the types of relationships, participants from 14 studies declared that they were living with their partner or in a committed relationship, married, in cohabitation, in a close relationship and relationship with partner. Levesque et al. (2014) have not specified their relationship.

Table 1 shows the characteristics of the 15 eligible studies. The journal Psychological Assessment had a higher impact factor, H index and better Qualis Capes. The access to the Journal Citation Reports database was through the Portal. Research on evidence of DCI validation was carried out in the continents of Europe, Asia, and America, mostly in Europe. The original German version of the DCI (Bodenmann, 2008) has been translated and adapted into 14 languages, English (Levesque et al., 2014; Randall et al., 2016), French and Italian (Ledermann et al., 2010), Hungarian (Martos et al., 2012), Spanish (Falconier et al., 2013), Portuguese – Portugal (Vedes et al., 2013), Romanian (Rusu et al., 2016), Chinese (Xu et al., 2016), Russian (Ekimchik & Kryukova, 2017), Persian (Fallahchai et al., 2019), Turkish (Kurt & Akbaş, 2019), Urdu (Shujja et al., 2020), Polish (Wendołowska et al., 2020), Tamil (Kanth et al., 2022). The German version by Gmelch et al. (2008) and Ledermann et al. (2010) were a validation study.

Discussion

This systematic review examined the evidence of validity of the internal structure of the Dyadic Coping Inventory (DCI). The findings confirmed the multidimensionality of the instrument. However, there were additive modifications in the factors structure, such as variations, focus on the problem and focus on emotion in SDC and CDC.

It is worth mentioning that, despite the research by Donato et al. (2009) having appeared recurrently in the records analyzed, it was not considered relevant for the objective of this review. This is because it did not analyze the properties of DCI, but rather of their predecessor, the FDTC-N. Falconier et al. (2019) was also not considered because it refers to an adaptation and validation of the DIC for financial stress, the Dyadic Coping Inventory for Financial Stress (DCIFS).

Levesque et al. (2014), Xu et al. (2016), and Kurt and Akbaş (2019) did not include the EDC factor. To Xu et al. (2016), this factor does not present theoretical support. Finally, although literature points out ramifications in NDC (Hostile, Ambivalent and Superficial), none of the studies found them (Falconier & Kuhn, 2019).

In short, most of the presented studies showcased substantial variation in the number of items, mainly leveraged by the removal of items in the SCO and SCP factor. The following are exceptions: Gmelch et. al. (2008), Vedes et al. (2013), Martos et al. (2012), Kurt and Akbaş (2019), and Fallahchai et al. (2019).

Reliability is an important psychometric property, since it concerns how well the instrument is able to reproduce similar results over time and space, considering different participants (Terwee et al., 2007). There are several techniques for measuring reliability, e.g., test-retest, split half (split test) and internal consistency (Murphy & Davidshofer, 2005). Among the studies analyzed, only Kurt and Akbaş (2019) used split half and Gmelch et al. (2008) the test-retest. The other studies used the internal consistency indicator, expressed by the Cronbach's Alpha coefficient.

There is no consensus in literature on the Cronbach's Alpha coefficient. However, Hair et al. (2019) highlighted those alpha values between .60 and .70 are considered the lower acceptability limit. In this sense, the total reliability index of the instrument pointed out in the studies ranged from .81 (Shujja et al., 2020) to .95 (Randall et al., 2016; Vedes et al., 2013). Although only Levesque et al. (2014), Xu et al. (2016), and Kurt and Akbaş (2019) have not reported full reliability since the creation of DCI by Bodenmann (2008), no study tested the existence of a global factor for Dyadic Coping. Therefore, although very prevalent in the studies presented here, the total reliability of the instrument seems to derive more from the sum and average reliability of the factor, which is empirically sustained.

Among the factors, the one that presented, with a higher regularity reliability, an index below α .60 is PSDC(O/P). This was verified in six studies (Falconier et al., 2013; Kanth et al., 2022; Randall et al., 2016; Rusu et al., 2016; Shujja et al., 2020; Xu et al., 2016).

Invariance measurements are related to evidence of validity (AERA, APA, & NCME, 2014). Of the fifteen studies, only seven presented invariance rates. This suggests a limitation of the studies. The indexes allow us to investigate how much the instrument seems to be adjusted to various orders of sample characteristics, e.g., genders, cultures, socio-educational levels. As for invariance measurements, the studies analyzed found at least partial scalar invariance among the tested groups (Falconier et al., 2013; Randall et al., 2016; Rusu et al., 2016; Shujja et al., 2020; Xu et al., 2016). In this respect, the research by Ledermann et al. (2010) which reached the level of configurational level between compared cultures.

Although the instrument undergoes modifications in its factorial structure and in the number of items, in general, it presents General Fit Index and factor weights that go from acceptable to very good. Based on the data of Invariance levels found, it makes it possible to allow adequate comparisons between genders and cultures. Finally, it is configured as an important tool to measure the quality of communication and the Dyadic Coping between gender and cultures.

By comparing groups, some studies made some mistakes in their analysis. Falconier et al. (2013) used a non-parametric test, Wilcoxon, to compare the means between men and women. This type of test is indicated to compare paired groups (Rey & Neuhäuser, 2011). Furthermore, the Measure of Central Tendency (MCT) in this case should be the Median. Thus, the authors should have opted for the Mann-Whitney test and the scores should have been indicated in ranks instead of means and standard deviations.

Other studies like Vedes et al. (2013) presents similar problems. Although the authors used the appropriate statistical test, the Mann-Whitney test for the non-normal unpaired samples, they used the mean as the MTC instead of using the Median.

Evidence of convergent, discriminant, concurrent or incremental validity was presented in all studies. However, these were not reported because they extrapolate the purpose of the review. To those interested in this information, see the studies in Table 1.

The International Test Comission (ITC, 2017) provides a set of guidelines for translating and adapting existing tests and for developing new tests. It highlights the differences between the terms of translation and adaptation of tests. The translation refers to a part of the process of adaptation that aims only at translating one language to another, that is, it is limited to preserve the meaning of the language and not consider the educational or psychological equivalence of the test. The adaptation has a broader and more complex procedure because it considers qualitative and quantitative evidence to preserve the linguistic, psychological and cultural differences of the target population.

In this review, psychometric studies of an existing instrument, DCI, were identified. Some authors reported only the process of translation and back-translation of the Spanish, Italian, French, Romanian, Portuguese versions (Falconier et al., 2013; Ledermann et al., 2010; Levesque et al., 2014; Randall et al., 2016; Rusu et al., 2016; Vedes et al., 2013). Fallahchai et al. (2019), the Persian version, includes a pilot study with qualified experts. Ekimchik and Kryukova (2017) and Kurt and Akbaş (2019), the Russian and Turkish versions respectively, include a pilot study with target population. Xu et al. (2016), the Chinese version, used recommendations only for translation procedures. Shujja et al. (2020), Wendołowska et al. (2020), and Kanth et al. (2022), the Urdu, Polish and Tamil versions, followed different guidelines to describe the translation and adaptation procedures. The Hungarian version did not mention the translation and adaptation procedures (Martos et al., 2012).

Although guidelines for test translation and adaptation have been available for approximately 20 years, it can still be observed in the literature that most test adaptations did not follow the guidelines. The ITC encourages the use of the guidelines as a primary source for disseminating best practices in research (ITC, 2017). Most studies carried out different reports of the instrument’s translation and adaptation procedures. This, perhaps, may compromise the analyses, inferences and results of the items, as well as the presentation of qualitative and quantitative evidence of the cross-cultural adaptation of the DIC. Future research could adopt a guide to reporting these procedures to favor replicability and promote evidence from cross-cultural research.

In this revision, the original German version of the DCI (Bodenmann, 2008) has been translated and adapted into 14 languages. The findings partially corroborate the number of languages in validation studies mentioned by Xu et al. (2016) and Shujja et al. (2020), German, Italian, French, Portuguese-Portugal, Romanian, Persian, English, Polish and Chinese versions, and five studies on the psychometric properties by Fallahchai et al. (2019), French, Portuguese-Portugal, English, Romanian and Chinese versions. Based on these data, it is suggested that research be carried out in other countries, including Brazil, to verify the psychometric properties of the DCI in different cultural contexts.

The Japanese DCI version by Yokotani and Kurosawa (2015) has been cited in other studies (Kanth et al., 2022; Shujja et al., 2020; Wendołowska et al., 2020). However, it was not considered in this review because it did not analyze the DCI psychometric properties. For Kanth et al. (2022), the DCI was translated to 24 languages, but they did not mention which ones.

The results of the fifteen studies that examined the DCI psychometric properties were across the American, European and Asian continents, Western and Eastern cultures. The difference between these cultures reveal specificities in the structure and dynamics of marital relationships (Kanth et al., 2022), which reflects differences in score between genders (Table 3). For example, in Pakistan, due to the practices of arranged, endogamous and exchange marriages and patriarchal society, women perceive their husbands as less oriented to share experience of stress with them and to incorporate assignments from their wives into their tasks (Shujja et al., 2020). In Polish, the family model is considered traditional and patriarchal (Wendołowska et al., 2020). So, women tend to seek help from their partners and assume the role of emotional support and care for the families, while men seek to help them, taking on their tasks and providing financial support to the family more regularly. This type of context is also seen in Latino families, as seen in Falconier et al. (2013) and in the Russian context (Ekimchik & Kryukova, 2017). In Chinese culture, the marital relationship is based on the culture of Confucianism and collectivism, which may explain the differences between the genders only in SC (Xu et al., 2016)

It should be noted that, despite data searches on Capes’ Portal de Periódicos and on Google Scholar search engine were made without refinement in the results, most of the studies were peer-reviewed. This suggests the possibility that non-peer-reviewed publications might have been disregarded in the review. On the other hand, peer-reviewed journals minimize biases and favor good quality of productions.

This review is considered original and unprecedented, as it compared the results of the DCI psychometric properties throughout its adaptation studies. The 15 studies presented two factorial structures and variation in the number of final items, either through PCA, EFA or CFA. The SC is fundamental in the analyses of DC behaviors. As additional strengths of this study, the following stand out: 1) presentation in Lingua Franca of studies in German, Hungarian, Russian and Turkish allowing for a greater reach of information; 2) the adoption of the terms factors, to refer to dimensions or subscales, and scales to refer to aggregated scales, perceptions or perspectives. This measure aimed to standardize the presentation of findings and reduce noise in understanding; 3) pointing out inaccuracies in the reports and in the methodological choices in the studies. During the selection of studies, a book was found that brings together various researches by the DCI in different cultures, for example, the psychometric properties in the Greek context. Due to the review eligibility criteria, only peer-reviewed journals and validity evidence based on the internal structure were included. It is recommended that subsequent similar reviews include other sources in their literature research. It’s also recommended that they check other evidence of validity based on external variables for a better range of results.