Religion and Social Identification:
Predictors for Healthcare Choices in Ghanaians in the Netherlands Onur Şahin
University of Amsterdam
Author Note
The author thanks Dr. Joseph Osafo for his input and guidance while starting this thesis project.
Furthermore, the author thanks Dr. Marieke Snijder for her help in executing Study 2. This thesis may be published in a scientific journal at a later moment in collaboration with HELIUS, but in its current form it is not a formal HELIUS publication that has been approved by the HELIUS board.
Thesis (Final version)
Supervisor: Prof. Dr. Bertjan Doosje Second supervisor: Dr. Michiel van Elk Student number: 10002734
Abstract
In Ghana, people often use traditional healthcare. To investigate whether religion, spirituality and social identification can predict the choices of people when choosing between healthcare providers, we conducted three studies with Ghanaians in the Netherlands. Interviews with five Ghanaians and an analysis of a large dataset (N = 2484) indicated that Ghanaians often pray alongside medical treatment, but religion or spirituality was not directly related to the choices they made. A quantitative within-design study with 41 Ghanaians in the Netherlands also found no predictive effect of religion or spirituality on intention to visit a doctor, traditional healer or a medical doctor with spiritual training, contrasting studies showing the importance of spirituality in healthcare in Ghana. Identification was a significant predictor of intention to visit a medical doctor with spiritual training. It is possible that the effect of identification was found because coexisting medical and spiritual explanations of illnesses results in
identification with the healer, who also provides both coexisting explanations. Religion and spirituality do not play a role in healthcare choices in Ghanaians in the Netherlands. However, Ghanaians seem to gain support and strength from private prayers for health.
Religion and Social Identification: Predictors for Healthcare Choices in Ghanaians in the Netherlands
Almost everyone in Ghana is religious. Around 71.2% percent of the population is Christian, 17.6% is Muslim, and the rest has a traditional indigenous religion or no religion (Ghana Statistical Service, 2010). These numbers indicate that religion is important to Ghanaians. For the purpose of this paper, religion is defined as the practices and beliefs of a religious organization (Shafranske & Maloney, 1990), whereas spirituality is defined as a personal relationship with a superior being (Richard & Bergin, 1997). One needs not to be religious to be spiritual, and one needs not to be spiritual to be religious. Especially people who have a high extrinsic religious orientation can be non-spiritual, since they practice their religion because it is a social norm which brings them protection, comfort and social status (Allport & Ross, 1967). People with an high degree of intrinsic religious orientation are more likely to use religion as a guide to their lives and as a personal experience, which is closer to spirituality. As religion is important in Ghana, so are the religious and spiritual leaders. People visit them for spiritual rituals or church sessions, but they also go to them for health issues. For example, people make use of herbal therapists, priests and priestesses and faith healers, also collectively defined as traditional medicine (Tabi, Powell & Hodnicki, 2006). An estimated 70% – 80% of Ghanaians use traditional medicine as their first resort for dealing with illnesses (Asamoah, Osafo & Agyapong, 2014). One reason for the high percentage of traditional healing in Ghana is that illnesses are often perceived in spiritual terms (Osafo, Agyapong & Asamoah, 2015). This perception makes that patients prefer a holistic approach that deals with both physical and spiritual issues, which is currently only offered by
traditional healers. This example shows that spirituality can be important in choices that people make in healthcare, however it is unclear whether it is decisive in the choices
Ghanaians make. In this paper, the importance of religion, spirituality and identification in healthcare choices will be investigated.
Besides spirituality, the availability and affordability of traditional healers seems to be an important factor that explains why many Ghanaians use their services. For example, in 2011 there were only 18 practicing psychiatrists, 19 psychologists, 1068 mental nurses and three psychiatric hospitals in Ghana (Roberts, Mogan & Asare, 2014). Sarpong (2008) noted that there was only one medical doctor for every 12.000 patients, whereas there were more than 45.000 traditional healers (approximately 1 for every 550 Ghanaians) and hundreds of prayer camps that provide healing services (Edwards, 2014).
Even though there are many traditional healers present, they have limitations.
Ghanaian health professionals reported in interviews that traditional healing “does not rely on accurate diagnosis” (White, 2015, p. 6), “sometimes neglects importance of dosage” (White, 2015, p. 6), and there have been reports about human right abuses in prayer camps, such as chaining and beating (Edwards, 2014; Read, Adiibokah & Nyame, 2009). These limitations of traditional healthcare in Ghana could be negated by collaborating with medical health
professionals, thus improving the overall healthcare. If the collaboration should succeed, it is of importance that patients who view illnesses in spiritual terms feel satisfied, as current medical healthcare does not satisfy the spiritual needs of Ghanaian patients (Koenig, 2008, as cited in Osafo, 2016). Some suggested collaborations that overcomes these disadvantages is training traditional healers in medical healthcare and training medical professionals in spirituality in health (Osafo, 2016). If medical doctors could address spiritual issues of patients when needed, patients who experience spiritual issues would be more likely to come to them. However, it is also important to investigate whether the lack of spirituality in medical healthcare is a reason for patients to avoid medical healthcare. Even though patients might
indicate that their spiritual needs are not fulfilled in medical healthcare, they might still prefer medical healthcare as it is over traditional healthcare or the collaborated forms of healthcare.
There seem to be many factors influencing whether people in Ghana would visit a traditional healer or a doctor. Since the availability of both healthcare providers seems to have a big influence, it would be interesting to see how Ghanaians view healthcare in a setting where doctors are common. This can be investigated in the Netherlands, which houses 23.000 inhabitants of Ghanaian descent and has an average distance of 1 km for a general practitioner visit (Central Bureau for Statistics, 2016). However, an investigation in the Netherlands introduces other factors that can influence the choices of Ghanaians. For example, language problems and lower levels of trust in the healthcare system were barriers in accessing healthcare for Ghanaians in the Netherlands (Boateng, Nicolaou, Dijkshoorn, Stronks & Agyemang, 2012). Knowledge of the health system, social support and awareness of diseases seemed to facilitate access to healthcare. Interestingly, although spirituality seems to be important in the choices that Ghanaians in Ghana make between traditional healers and doctors, the role of spirituality has not been investigated in Ghanaians living elsewhere. Also, the studies that have been discussed so far are all qualitative studies. Even though these studies are informative, relationships between variables cannot be investigated and the level of generalization is limited. To fill this gap, we propose a study that uses mixed methods. This way, we will better able to generalize our findings.
One of the important aspects of traditional healers is that they fall in line with the spiritual beliefs of patients (Ae-Ngibise et al., 2010; Tabi et al., 2006). This is not only the case for Ghanaians; there are studies showing that people across cultures use both natural and supernatural explanations to explain events, such as illnesses (Busch, Watson-Jones &
Legare, 2016). For example, people from India and the United States explained symptoms of illnesses predominantly by using a biological model, but the same individuals also endorsed
supernatural explanations (Raman & Gelman, 2004). Although religion seems to be important in coexisting explanations, spirituality might be more important than religion since any form of “supernatural” explanations are included (Legare & Visala, 2011). Training doctors in spiritual issues, as suggested by Osafo (2016) and Asante and Raphael (2013), might help with understanding and dealing with the coexisting explanations of patients. Patients with coexisting explanations of illnesses will likely prefer these healthcare providers over solely doctors or traditional healers, since they are able to address both biological and spiritual issues. It is also possible that religious/spiritual patients prefer traditional healers or doctors with spiritual training because they identify more with them. A theory that can help
understand this process of identification is the self-categorization theory (Turner, Hogg, Oakes, Reicher & Wetherell, 1987). According to this theory, groups can be categorized at at least three hierarchical levels of categorization. The superordinate level is the highest level, for example the category of “human”. The intermediate level defines social groups, for example “healthcare provider”. The subordinate level is based on differentiations between individuals, defining one as a specific individual. Recategorization has been shown to be effective at manipulating categories and affects how people categorize others as being members of the same or different group (Crisp, Stone & Hall, 2006; Guerra, Rebelo, Monteiro & Gaertner, 2013).
For a Ghanaian patient, traditional healers and doctors are both healers at an intermediate level of abstraction. At the subordinate level of abstraction the differences become salient: Traditional healers use herbs or spiritual means to heal, whereas doctors use medicine or psychotherapy. However, if some doctors are also trained in spirituality and would be able to address spiritual problems, they would form a third category.
If a person has the choice between the three categories of traditional healers, doctors and doctors with spiritual training (MIX), they likely perceive similarities between the
categories and themselves and have a certain degree of identification with each of them. Previous studies showed that perceiving similarities between the self and other, or
identification, leads to ingroup-favoritism. This is a greater liking for the similar other than the dissimilar other (Robbins & Krueger, 2005). Therefore, we expect that identification is also important in the choices Ghanaians make between the categories. We argue that when traditional healers, doctors and MIX are presented as options for treating an illness,
Ghanaians who are more religious or spiritual will show greater intentions of visiting traditional healers or MIX than visiting doctors. This effect will likely be mediated by the level of identification that spiritual people have with healers that can address spiritual issues. Ghanaians that are less spiritual will be more likely to have greater intentions to visit doctors and MIX than visiting traditional healers, since they want spiritual healing and also identify less with traditional healers. Therefore, MIX will be a middle ground for spiritual and less spiritual people in terms of popularity.
The key purpose of the current studies will be to investigate how religion, spirituality and identification is related to preferences of Ghanaians (living in the Netherlands) in
healthcare. The hypotheses are:
1 Ghanaian participants will have a greater intention of visiting MIX and traditional healers than visiting doctors when ill.
2 Religion and spirituality will positively predict the intention to visit traditional healers, but negatively predict the intention to visit doctors. Religion and spirituality will not predict intention to visit MIX, since this will be a viable option for everyone. 3 Identification with the healers will mediate the relationship between
religion/spirituality and the intention to visit the doctors and traditional healers, since the participants will likely identify with the healer based on whether they offer a religious/spiritual perspective to their problems.
These hypotheses will be tested using three separate studies. The first study will consist of interviews. This study is meant to get an impression of how Ghanaians in the Netherlands view traditional and medical healthcare in the Netherlands and to identify which factors might play a role in choosing between healers. In the second study, we analyze an existing dataset, in which we investigate whether religion and acculturation play roles in visiting medical specialists, traditional healers and psychosocial healthcare providers. The third study will be a quantitative study in which we investigate the relationship of religion, spirituality and
identification with the intention to visit a doctor, traditional healer and MIX (using a within-design with three vignettes).
Study 1
To explore how Ghanaians in the Netherlands think about the different forms of healthcare, semi-structured interviews were conducted. Questions were asked to assess how Ghanaians interpret illnesses and what influences their choices of going to a specific healer. Using questions that could be interpreted broadly and in different ways, it was ascertained that participants were not steered into giving religious/spiritual or biological explanations for illnesses. For example, when asking what causes mental or physical illnesses, no follow up question was asked about physical or spiritual possibilities of causes to avoid steering the participant into certain directions.
Participants
Community centers and churches in The Hague and Amsterdam were approached to recruit participants with a Ghanaian background. There were no criteria for exclusion. Five participants were interviewed (3 males, mean age = 51.20, SD = 14.89). All participants indicated being Christian. These interviews were conducted in community centers or the houses of the participants. They received €10 as a reward.
The interview guideline that was used can be found in Appendix A. Broadly speaking, we covered the following topics: Experience in medical and traditional healthcare in Ghana and the Netherlands, enablers and barriers of healthcare, importance of religion when choosing a healthcare provider, importance of religion in the healing process, and perceived causes of illnesses.
Results
The interviews were analyzed according to interpretive phenomenological analysis (IPA) (Smith & Osborn, 2015). Using this method, the meaning that experiences, events or situations hold to the participants can be analyzed. This method was considered useful in this study, since religious/spiritual meaning was often assigned to events and experiences in Ghanaians living in Ghana.
The participants gave similar answers to the questions, despite being interviewed in different settings in different cities. The themes that were uncovered were: 1. Satisfaction with healthcare system, 2. Traditional beliefs and age, 3. Integration in the Netherlands and 4. Coexisting explanations. These themes played a central role in all five interviews.
Satisfaction with healthcare system. This theme is important since it includes
information about how Ghanaians view the medical healthcare that is provided in the Netherlands and the healthcare that is provided in Ghana. Most participants indicated satisfaction with the Dutch healthcare system. This is their personal experience, but also the experience from people in their network:
Participant 1: “My personal experience with them is that most of them really like it here.” Participant 3: “Somebody says “The (Dutch) system is not good”. Somebody says “the system is good”. But for me, it’s good for me.”
Participant 4: “For myself, I’m satisfied. Because the way they (Dutch healthcare) treat me” Participant 5: “In my opinion, everyone likes it (Dutch healthcare).”
As participant 3 indicated, there are also Ghanaians that do not trust the healthcare system in the Netherlands. The reason is best explained by the participants:
Participant 2: “60% of people don’t trust it here.”
Participant 4: “Some think that they don’t have the right medicine because they think that they’re not Nederlands (Dutch), so they don’t take good care of them. For foreigners, they give less expensive medicine.”
Some Ghanaians seem to distrust their doctors, thinking that they get medicine from a lower quality compared to Dutch people. Overall, Ghanaians seem to be pleased with the Dutch healthcare system. The big advantage seems to be the accessibility, thanks to healthcare insurance. The lack of insurance in Ghana seems to make treatment expensive there, which results in people searching for other, cheaper means to get treatment for their problems: Participant 1: “Because in Ghana, the insurance, we do not have the proper healthcare insurance system. It’s really costly.”
Participant 4:
Question: Yeah, that is, people use herbal to care for themselves. Response: Because that’s cheaper.
Q: Yeah, that’s cheaper then. Participant 5:
Q: So how common is traditional healing in the Ghanaian community?
R: It’s a lot, because when you are poor, that is where you’re supposed to go.
The participants mostly seem satisfied with the Dutch healthcare system. In Ghana, the problem seems to be how expensive it is, which is a reason for people to use alternative means of healing, like traditional medicine. People in the Netherlands are insured, so money seems to play a minor role in the choices between traditional and medical healthcare.
Traditional beliefs seem to play a more important role in using traditional medicine in the Netherlands.
Traditional beliefs and age. Having traditional beliefs is important in understanding
the usage of traditional medicine. Although people have access to medical healthcare in the Netherlands, some Ghanaians believe traditional medicine is also good or even better for them compared to medical healthcare. The participants suggest that especially older Ghanaians still hold tightly to their traditional beliefs:
Participant 1: “Yeah the older generation here, they, they sometimes like to go back and use traditional… When they go on holidays they bring traditional medication.”
Participant 2:
R: But in the older group, 80% like the traditional one. Because during my mom and my dad’s time there were only few doctors so they used only the traditional ones.
Q: Okay, because there was not much other choice.
R: Even they go to the hospital. They still have the traditional medicine at home, taking it. Q: They go to a hospital but, both, they do both.
R: But they prefer the traditional one than the hospital.
Q: And why do you think that is changing in the younger generation?
R: That’s because of education. Here in Holland there is no traditional, so you have no choice. So I’m used going to the hospital, even when I’m in Ghana and I got sick. Okay, let me go to the doctor because I’m used to that.
Participant 3 is 73 years old himself and likes using traditional medicine:
Participant 3: “You go to hospital, take your pills, take your medicine. I take my African traditional healing. No problem.”
Participant 3: “The important thing is, you have to go to a doctor. Traditional healer is, they cannot understand. You go to doctor’s office and you know the sickness before you go to the traditional healer. Maybe then, I say that I’m a diabetes patient. I know that I’m a diabetes patient and the doctor gives me medicine. So if I go to Ghana now, I go, I’ll ask, “I have this sickness, I need African medicine, traditional medicine”, I can get. And I can take both.” Participant 4:
R: Older generation, they understand everything, because they have done it before. But the youngsters, I don’t think they like that way. They want to, they want doctor for healing. Q: Because the youngsters, they don’t understand traditional healing.
R: How traditional healing works, yes.
Q: And this accounts for both herbal medicine and spiritual healing?
R: Yes, both is working for them, those who are not rich. Sometimes it works for them. The participants indicate that sometimes herbs can be bought in Ghanaians shops in the Netherlands, but some, like participant 3, imports the medicine from Ghana. This shows some people use traditional medicine in the Netherlands, but mostly elderly Ghanaians.
Integration in the Netherlands. One of the reasons that some Ghanaians do not like
the doctors in the Netherlands seems to be an issue of integration. If Ghanaians do not speak Dutch or English, this can be a problem in communicating with the caretaker, as indicated earlier (Boateng et al., 2012). However, this difficulty in communication does not seem to lead people into using traditional healing, rather they try to solve this problem with other means:
Participant 1:
Q: So is that the reason why people go to traditional healers? R: No, language will not be.
Q: And why is that? Is it because people will find a way, like you said, they will bring someone who translates?
R: In the case of the housedoctor, in that case, if the person does not have the knowledge, that’s okay. I could take somebody with me and it will help. (…) But language will not drive them to go and seek a traditional healer. It is only when the orthodox (i.e. medical) medicine has failed them. That is when they go.
Participant 5 indicates a the same thing when asked whether language is a problem:
Participants 5:“Yeah, little bit, but they always have to have somebody with them. Translator or something.”
Participant 3 indicates that there are some Ghanaian doctors working in hospitals in Amsterdam, which makes it easier for those that cannot communicate:
Participant 3: “So sometimes if you go and you cannot express yourself, then they ask you “where are you from, are you from Ghana”? Then they start speaking your language.” Participant 4: “For many people, because they don’t speak the language good. So, it’s very, sometimes very difficult for them to explain themselves.”
He furthermore indicates he thinks the lack of communication could have led to misunderstanding the illness:
Participant 4: “Some people cannot explain themselves for their sickness, so they (the doctors) give them wrong medicine”
Language does indeed seem to be problem in Ghanaians in the Netherlands, however bringing someone who can translate seems to solve the problem. Due to this solution, lack of
knowledge of Dutch or English does not seem to result in visits to traditional healers.
Coexisting explanations. The participants indicated that they do pray to God for
specifically aimed to cure their illness. Usually, they combine taking medicine and praying (both alone and in the church) as indicated in the following citations:
Participant 1: “People don’t do a solo treating. They mostly go to a housedoctor (general practitioner) and still pray for the condition.”
Participant 4:
Q: Okay, let’s see. If people get sick, do they go to medical doctor? Or do they only go to pray?
R: No, they go to medical doctor, they take the medicine, they do everything normal, what the doctor prescribes. But apart from that, we go to church and pray that the medicine they took will work for them and they can get better.
Participant 2 also indicated that there is a group of people that only prays to God for healing: Q: How much does religion do you think play a role in illnesses, diseases?
R: Plays a major role. Because we think, when you go to church, you pray when you’re sick. So it is a major influence. I know some even religious groups where the sick don’t even have to go to the hospital.
Q: Oh they just…
R: Go to the church and pray, they’ll be healed. Q: They never go to the hospital?
R: They don’t want to. Q: They don’t want to?
R: Yeah because of religious bases. They think God will heal them.
The participants pray to God for healing, however it never replaces medical treatment. It seems that it is more of a complementary approach rather than focus on one treatment exclusively, which seems to be a form of coexisting explanations of illnesses. Participants also indicated no preference for a Christian doctor over a doctor with other beliefs. There is a
group of Ghanaians that do focus on solely seeking treatment from God, this seems to be a minority group as only one participant mentioned some people doing this. When it comes to causes of illnesses, all participants assigned psychologic or biologic causes.
Participant 4:
Q: So what do you think causes mental illnesses? So for example schizophrenia, sometimes people hear voices.
R: It’s stress.
I: And physical illnesses?
R: Physical illness is, overload, if you work too much, the body cannot contain it. Participant 3:
Q: What do you think the cause is of mental illness?
B: For that one I don’t know. Maybe it is from parents, sometimes it’s from the father, sometimes it’s from the mother. The sickness, they have. Sometimes it comes from the family of the mother. Sometimes it comes from the family of the father. Sometimes, yeah.
I: So you mean genetic? It’s passed on? B: Yeah, yeah.
This seems to contrast with research in Ghana, where people indicated that some illnesses can be caused by spiritual issues or that God causes these illnesses (Osafo, Agyapong &
Asamoah, 2015). Since illnesses are not viewed in spiritual terms by Ghanaians in the
Netherlands, it is likely that they will not look for healers who view illnesses in spiritual terms either.
Discussion
The purpose of this first study was to familiarize ourselves with the Ghanaian
community in the Netherlands, especially on topics related to healthcare. It must be noted that only five participants took part in this study, so the options for generalization are limited.
However, often participants had similar answers to the questions, which increases the credibility and likelihood that their answers reflect the ideas and attitudes of the Ghanaian community.
People seem to use traditional healing for two major reasons. Insurance is not common which makes medical healthcare expensive in Ghana. People thus resort to
traditional healthcare. In the Netherlands, medical healthcare is available and satisfactory for most Ghanaians. This finding is in line with previous research showing that traditional
healthcare in Ghana is overrepresented and more accessible than medical healthcare (Sarpong, 2008). Another reason seems to be belief in the working of traditional medicine. The two older participants (2 and 3) indicated using traditional medicine because it works for them. The other participants also indicate that older people indeed use traditional medicine, because it is what they are used to. Issues related to integration, i.e. language problems, seem to be present. However, these issues do not seem to be related to healthcare choices. Lastly, Ghanaians seem to pray for healing alongside medical healthcare. Because they often pray alongside medical treatment, they seem to have coexisting explanations for illnesses. Therefore, Ghanaians might have a preference for a MIX rather than a doctor or traditional healer alone. However, when asked about the causes of illnesses, they did not report
supernatural causes, which makes the choice for MIX or traditional healers for treatment less likely. It is possible that the causes are seen as biological, but as an underlying force they perceive God. Another possibility is that they use spiritual means to treat illnesses, which they perceive as purely biological. In all, the interviewees mostly seem to prefer doctors over traditional healing methods, even though all of them are religious. Traditional Ghanaian beliefs in elderly Ghanaians seem to be an important reason why some Ghanaians in the Netherlands still use traditional healthcare.
Study 2
There are many possible factors that could predict whether people would go to a traditional healer or a medical doctor. During the interviews, the factors religion, budget and traditional beliefs were investigated. This study was limited to five participants, which helps in getting an impression of the Ghanaian community in the Netherlands, but lacks in sample size to make inferences about the population.
Therefore, a dataset that was made available by HELIUS (Stronks et al., 2013) was analyzed to investigate whether religion and acculturation were predictors of visits to medical and traditional. The dataset includes participants of Dutch, Turkish, Moroccan, Surinamese and Ghanaian ethnicity in the Netherlands, which makes a comparison between groups possible. This is useful to investigate whether Ghanaians are more religious and whether they use traditional healthcare more than other ethnic groups. We hypothesized that Ghanaians would have a greater intention to visit traditional healers and doctors with spiritual training than visiting doctors. This dataset has data on visits to traditional healers and doctors by Ghanaians and their religiosity, thus it can give some insight regarding the hypotheses.
It is also possible that acculturation makes a difference, since the interviews showed that traditional beliefs were important for those who used traditional medicine. Ghanaians in the Netherlands who are attached to their own culture and did not adopt to the host Dutch culture (defined as ‘separated’ by Berry, 1997) are possibly similar in traditional beliefs to Ghanaians in Ghana. Ghanaians who adapted to the culture of the Netherlands (defined as integrated, or, if the culture of origin is rejected, as assimilated by Berry, 1997) should have less traditional beliefs than Ghanaians in Ghana. Therefore, separated Ghanaians will likely have visited traditional healers more often than integrated or assimilated Ghanaians have. Furthermore, previous studies have also shown that rejecting both the host culture and culture of origin, or marginalization (Berry, 1997), is a risk factor for poor health (Lynam & Cowley,
2007). This could result into more visits to healthcare providers by marginalized Ghanaians when compared to integrated Ghanaians. Therefore, the current study will investigate the role of religion and acculturation in visits to healthcare providers in the past year.
Participants
The participants were part of a random sample of the municipality registry of Amsterdam, stratified by ethnicity (see Stronks et al. (2013) for more information). The requested dataset included 4641 Dutch participants (mean age = 46.14, SD = 14.05, 45.9% male), 8363 Surinamese participants (mean age = 46.56, SD = 13.12, 42.9% male), 4067 Turkish participants (mean age = 39.90, SD = 12.45, 45.1% male), 4337 Moroccan
participants (mean age = 39.73, SD = 13.05, 37.8% male) and 2484 Ghanaian participants (mean age = 44.24, SD = 11.51, 38.6% male) – all living in Amsterdam.
Materials
As the study was exploratory, variables that were measured by the HELIUS questionnaire were requested from their research team (Stronks et al., 2013). This
questionnaire is available upon request. The variables that were used for this study will be discussed in this section.
Demographical factors. HELIUS registered standard demographical information,
like age, gender (male/female), education level (four categories: no, lower, intermediate and higher education), occupation level (five categories: elementary, lower, intermediate, higher and academic occupations) and residence duration in the Netherlands (in years).
Health. The dataset contained information about the physical and mental health state
of the participants. This was measured using the SF-12 questionnaire (Ware, Kosinski & Keller, 1996). The subscales Physical Component Summary (PCS) and Mental Component Summary (MCS) were included to include physical and mental wellbeing in the analyses
(test-retest reliability = 0.89 and 0.76 in the US and UK, respectively). The scores can range from 0 – 100, a higher score indicates better health.
Acculturation. As traditional beliefs seemed to be important for the interviewed
Ghanaians, measures of acculturation were included in the analyses. In this section, the measures are briefly discussed. More extensive information about the measures are available in Appendix B.
Ethnic identity was measured by asking the participants to indicate how much they felt Dutch and how much they felt Ghanaian. HELIUS categorized the participants into four categories, using the acculturation strategies as defined by Berry (1997). These strategies are integration, assimilation, separation and marginalization.
Cultural orientation was measured by using an adjusted version of the Psychological Acculturation Scale (PAS) (Stevens et al., 2004). Participants were categorized into the four acculturation strategies. Cronbach’s α of the PAS was .85 in a Dutch sample and .73 in a Moroccan sample.
Lastly, the cultural diversity of the network was used in the analyses. Participants had answered questions about whether they had friends from their own ethnicity or the Dutch and how much contact they had with both groups. This way, whether the network of the
participants consisted of mainly Ghanaians, Dutch, neither or both was measured. Again, HELIUS categorized the participants into the four acculturation strategies.
Religion. To investigate the role of religion, two questions from the HELIUS
questionnaire were used in the analyses. One question assessed whether the participants was religious (yes/no) and how often the participants visited a house of worship. This was coded in five categories: never, less than once a month, once a month, once every two weeks and at least once a week.
Satisfaction with healthcare. To assess whether the participants were satisfied with
their current healthcare, a question asking the participants whether they were satisfied with their current GP was used. This ranged from 1 (dissatisfied) – 5 (very satisfied).
Visits to healthcare providers. As the dependent variable, participants were given a list of
medical specialists, traditional healers and psychosocial healthcare providers. They had to indicate whether they visited one of these healthcare providers in the past 12 months. If they visited one of the healthcare providers listed, their answer was recorded as a “yes” in the respective category. For example, if they visited a cardiologist in the past 12 months, it was coded as a “yes” in the category of medical specialists, if they visited a herbal therapist, it was coded as a “yes” in the category of traditional healers. Thus, three categories of healthcare providers were categorically coded as “yes” (visited) or “no” (did not visit). Participant who left the answer blank were left out of the analyses.
Procedure
Participants were invited to participate in the HELIUS study by mail. They could indicate a preference to fill in the questionnaire online or on paper. If needed, a trained, ethnically matched interviewer could offer assistance.
Results
All analyses were conducted using SPSS (version 23). Descriptive statistics were used to summarize the most important variables per ethnic group. Since sample size differed per ethnic group and per question (due to missing values per question), percentages were used to make comparison between the ethnic groups easier. These descriptives of religion and acculturation can be found in Table 1, in Table 2 the visits to healthcare providers per ethnic group can be viewed
The most interesting finding between ethnic groups is that there is a big difference between visits to houses of worship, X²(20) = 2891.323, p < .001. Around 75.6% of
Ghanaians indicated visiting a house of worship at least once a week, whereas this percentage is at least 37 percentage points lower in the other ethnic groups. This is in line with previous studies, which shows that religion is important for Ghanaians. An interesting oddity is that there is a difference in visits to medical specialists between the ethnic groups, X²(6) = 217.589, p < .001. Ghanaians used medical specialists less than other ethnic groups. A regression analysis might unveil whether religion plays a role in these findings.
Since this is an exploratory study, backward elimination logistic regression analyses were conducted to predict whether participants visited the healers in the past year. Because initial frequency analyses revealed that the traditional healer, spiritual/religious guide, natural healer and paranormal healer were visited quite infrequently, these variables were combined into one variable called traditional healers. Therefore, three separate logistic regressions were conducted to predict visitations of medical specialists, traditional healers and psychosocial healers. The independent variables were: age, gender, education level, occupation level, residence duration in the Netherlands, PCS, MCS, ethnic identity, cultural orientation, cultural network, currently religious, frequency of visiting house of worship, and satisfaction with current GP. Considering the scope of the current study, logistic regressions were only conducted for Ghanaians.
A logistic regression predicting visits to at least one medical specialist in the past year by Ghanaians was conducted. Differences between categories of categorical variables were investigated with a Wald test. Due to missing data on some variables, 1359 participants were included in the analyses. The final model, after backward elimination, showed that gender, education level, PCS, MCS, residence duration, cultural identity and cultural network were significant predictors of visitations of medical specialists in the past year (see Table 3). Women indicated to have visited medical specialist 1.8 times more often than men did. Furthermore, Ghanaians with an intermediate or higher education level made more use of
medical specialists than those without education, b = .549, Wald X²(1) = 10.603, p = .001 and b = 1.032, Wald X²(1) = 16.478, p < .001, respectively. Physical and mental health also predict, unsurprisingly, visitations to medical specialists. The odd ratios indicate that Ghanaians with better health have a smaller chance of having visited a medical specialist. Furthermore, a longer residence duration in the Netherlands seems to increase the chance that someone has visited a specialist. Ghanaians with a marginalized identity seemed to have visited specialists more than 4 times as often than those with an integrated identity, b = 1.435, Wald X²(1) = 6.685, p = .010. An ANOVA was conducted to investigate physical health differed between ethnic groups, as a lower physical health in Ghanaians with a marginalized identity might have led to more visits to medical specialists. The multivariate test showed that there was a main effect of ethnic identity on physical health, F(6,37270) = 28.320, p < .001. However, Bonferroni-corrected post hoc tests showed that Ghanaians with a marginalized identity (M = 46.190, SE = .519) did not differ in physical health than those with an integrated identity (M = 47.372, SE = .083), p = .147. Lastly, Ghanaians with a separated network seem to have visited specialists less often than Ghanaians with an integrated network, b = -.512, Wald X²(1) = 13.456, p < .001.
Table 1
Percentages of Categories of Acculturation in Ghanaians in the Netherlands and Religiosity of all Ethnic Groups.
Questions Categories Ghanai
an
Dutch Surinamese Turkish Moroccan
Ethnic identity Integration Assimilation Separation Marginalization 70.3 1.2 26.9 1.6 Cultural orientation Integration
Assimilation Separation Marginalization 73.9 1.1 23.9 1.1 Cultural network Integration
Assimilation Separation Marginalization 28.8 3.9 55.8 11.5
Currently religious Yes* 88.9 13.2 77.6 93.4 98.2
Visited house of worship
Never*
Less than once a month* Once a month*
Twice a month* At least once a week*
3.9 6 5.6 8.9 75.6 39 33.9 6.3 8.4 12.5 36.7 30.3 10.3 5.2 17.6 22.4 21 10.6 7.4 38.6 29.4 22 8.1 6.8 33.6 * indicates a difference between the ethnic groups, p < .001
Table 2
Percentage of Visits to Healthcare Providers in the Past Year per Ethnic Group
Category of healer Ghanaians Dutch Surinamese Turkish Moroccan
Paranormal healer 0.2 1.1 1.2 0.1 0.2 Spiritual guide 3.2 0.3 1.9 0.9 1.7 Traditional healer 1.2 0.3 0.8 0.6 0.7 Natural healer 1.2 1.6 0.8 0.7 0.7 Medical specialist 42.7 50.8 58 55.4 50 Psychosocial healer 15.5 15.8 18.7 18.2 18.3
Table 3
Beta Values, Standard Errors and Odd Ratios (with 95% Confidence Interval) of the Significant Logistic Regression Model, Dependent Variable is Having Visited a Medical Specialist in the Past Year.
Included variable B (SE) OR (95% CI)
Gender*** 0.59 (0.13) 1.80 (1.34 – 2.32) Education (no)a Education (lower) 0.13 (0.15) 1.14 (0.85 – 1.52) Education (intermediate)** 0.55 (0.17) 1.73 (1.24 – 2.41) Education (higher)*** 1.03 (0.25) 2.80 (1.71- 4.62) PCS*** -0.05 (0.01) 0.96 (0.94 – 0.97) MCS*** -0.03 (0.01) 0.97 (0.96 – 0.99) Residence duration** 0.03 (0.01) 1.03 (1.01 – 1.04) Cultural identity (integration)a
Cultural identity (assimilation) -0.55 (0.65) 0.58 (0.16 – 2.08) Cultural identity (separation) -0.1 (0.14) 0.99 (0.76 – 1.29) Cultural identity (marginalization)* 1.44 (0.56) 4.20 (1.42 – 12.47) Cultural network (integration)a
Cultural network (assimilation) 0.75 (0.39) 2.12 (0.98 – 4.55) Cultural network (separation)*** -0.51 (0.14) 0.60 (0.46 – 0.79) Cultural network (marginalization) -0.12 (0.20) 0.87 (0.60 – 1.31)
Constant 2.31 (0.50) 10.09
*** p < .001 ** p < .01 * p < .05 a: reference group
Note: PCS = Physical Component Score, measuring physical health, MCS = Mental Component score, measuring mental health Model X²(13) = 169.15, p < .001, R² = .12 (Cox & Snell), .16 (Nagelkerke)
Another stepwise backward logistic regression was conducted to predict the visits to traditional healers by Ghanaians (Table 4). A total of 1349 participants were included in the analyses. The final model showed that gender, physical health and mental health predicted visits to traditional healers by Ghanaians, where women were twice as likely to have visited a traditional/spiritual healer in the past year. A better health predicted fewer visits to traditional healers.
Table 4
Beta Values, Standard Errors and Odd Ratios (with 95% Confidence Interval) of the Significant Logistic Regression Model, Dependent Variable is Having Visited a Traditional Healer in the Past Year.
Included variable B (SE) OR (95% CI)
Gender* 0.699 (0.30) 2.01 (1.11 – 3.64)
PCS** -0.04 (0.01) 0.96 (0.94 – 0.99)
MCS** -0.03 (0.01) 0.97 (0.95 – 0.99)
Constant -0.10 (0.82) 0.91
*** p < .001 ** p < .01 * p < .05 Note: PCS = Physical Component Score, measuring physical health, MCS = Mental Component score, measuring mental health Model X²(13) = 27.44, p < .001, R² = .02 (Cox & Snell), .06 (Nagelkerke)
Lastly, a stepwise backward logistic regression was conducted to predict the visits to psychosocial healthcare providers by Ghanaians. A total of 1347 participants were included in the analyses (see Table 5).
Significant predictors were physical and mental health. Satisfaction also predicted visits to psychosocial caretakers, Wald tests indicated that Ghanaians that were moderately,
reasonably and clearly satisfied with their GP visited psychosocial healthcare providers less often than Ghanaians who were dissatisfied (b = 1.517, Wald X²(1) = 5.409, p = .020, b = -1.193, Wald X²(1) = 5.685, p = .017, b = -1.172, Wald X²(1) = 5.821, p = .016, respectively).
Table 5
Beta Values, Standard Errors and Odd Ratios (with 95% Confidence Interval) of the
Significant Logistic Regression Model, Dependent Variable is Having Visited a Psychosocial Healthcare Provider in the Past Year.
Included variable B (SE) OR (95% CI)
PCS*** -0.05 (0.01) 0.95 (0.94 – 0.97)
MCS*** -0.07 (0.01) 0.94 (0.92 – 0.95)
Cultural identity (integration)a
Cultural identity (assimilation) -19.58 (10515.54)
0
Cultural identity (separation) -0.03 (0.20) 0.98 (0.66 – 1.44) Cultural identity (marginalization) 0.93 (0.53) 2.54 (0.90 – 7.19) Cultural network (integration)a
Cultural network (assimilation) 0.72 (0.43) 2.05 (0.88 – 4.80) Cultural network (separation) -0.32 (0.19) 0.73 (0.50 – 1.06) Cultural network (marginalization) -0.01 (0.28) 0.99 (0.58 – 1.70) Satisfaction with GP (dissatisfied)a
Satisfaction with GP (moderately satis.)* -1.52 (0.65) 0.22 (0.06 – 0.79) Satisfaction with GP (reasonably satis.)* -1.20 (0.50) 0.30 (0.11 – 0.81) Satisfaction with GP (clearly satisfied)* -1.17 (0.49) 0.31 (0.12 – 0.80) Satisfaction with GP (very satisfied) -0.80 (0.48) 0.45 (0.18 – 1.15)
Constant 4.74 (0.71) 114.92 *** p < .001 ** p < .01 * p < .05 a: reference group
Note: PCS = Physical Component Score, measuring physical health, MCS = Mental Component score, measuring mental health
Model X²(12) = 141.33, p < .001, R² = .10 (Cox & Snell), .18 (Nagelkerke)
The same models ran with forward stepwise regression instead of backward regression
yielded similar results. Only the third model, with psychosocial caretakers, differed in that the non-significant predictors were not in the model.
Discussion
This study confirms that Ghanaians are more religious than other ethnic groups, because most Ghanaians visit church at least once a week. To investigate whether religion and acculturation could predict visitations to medical specialists, traditional healers and psychosocial healthcare providers, three separate logistic regression analyses were conducted. The most important finding is that religion was not a significant predictor of visits to any healthcare provider. Since a great majority of Ghanaians indicated that they were religious and have visited a house of worship at least once a week, it is likely that due to ceiling effects (thus lack of variability) these factors could not predict anything. In the analyses it became clear that women use medical specialists and traditional healers more often than men. For traditional healers, gender was even the only significant predictor, apart from one’s own health situation predicting health seeking behavior. These findings seem in line with previous studies that showed that women utilize healthcare more often than men do (Bertakis, Azari, Helms, Callahan & Robbins, 2000). Furthermore, a higher education level seemed to predict more visits to medical specialists. It is possible those who have a higher education level can more easily find their way in the healthcare system. The number of years of residence in the
Netherlands is also positively predicting visits to medical specialists. It is possible that as people have stayed longer in the Netherlands, they are able to find their way to a medical specialist easier than Ghanaian who are new immigrants. One’s physical and mental health are significant predictors in all cases. A better health results in a smaller chance of having visited one of the healthcare providers.
More importantly, having a separated, compared to integrated, network negatively predicts visits to medical specialists. A possible explanation is that people who have an integrated network have easier access to the healthcare system, as Dutch friends can refer or help with finding ones way in the healthcare system. In contrast, having a mostly Ghanaian network would make it more difficult to understand the Dutch healthcare system if most friends are also migrants. Furthermore, being marginalized, compared to integrated, in identity positively predicts visitations of medical specialists. This is in line with previous studies that have shown that marginalized people experience more health problems than integrated people (Lynam & Cowley, 2007). However, Ghanaians with a marginalized identity did not indicate having a worse physical health than those with an integrated identity. It is possible that the few Ghanaians with a marginalized identity went to medical specialists more often by chance, which has an effect simply because those with a marginalized identity are few. Lastly, the negative predictive effect of satisfaction with GP on visiting psychosocial healthcare providers is likely due to the GP handling psychosocial issues if the relationship with the patient is good.
The predictive effect of religion on healthcare choices could not be fully investigated in the current study, mainly due to limitations in design and the lack of variety in answers. There does seem to be a cultural link, which is present while predicting visits to medical specialists. The next study will be an improvement over Studies 1 and 2, by focusing more on
religion and identification with the healthcare providers in Ghanaians using an experimental design and a more complex measurement of religion and spirituality.
Study 3
In both Study 1 and Study 2, the conclusions were that religion did not predict the choices that Ghanaians make in healthcare. However, both studies had its limitation, one being a qualitative study with few respondents and the other, despite having a big sample size, having limitations in questions. The current quantitative study introduces a more complex measurement of religiosity to allow more variability in the answers. Furthermore, the role of identification with the care provider could not be investigated in the previous studies. The reasoning behind the idea of identification is that if Ghanaians are confronted with healers that belong to different categories, they will prefer the healer that has the most in common with themselves. Thus, they identify more with a healthcare provider who is also described in terms of spirituality if they are more spiritual themselves. As indicated from the previous studies, religion is important to Ghanaians. Therefore, it is hypothesized that Ghanaians will have a greater intention to visit MIX and traditional healers than to visit doctors. Hypothesis 2 indicated that religion and spirituality will positively predict the intention to visit traditional healers, but negatively predict the intention to visit doctors. No prediction of religion or spirituality for intention to visit MIX is expected, since they are a viable option for all participants. Religiosity and spirituality should also lead to a higher identification with healers who are also religious or spiritual. Therefore, it is expected that identification will mediate the relationship between religion/spirituality and the intention to visit the doctors or traditional healers.
To test these hypotheses, the current study has a quantitative within-design with three types of healthcare providers as conditions. The difference with Study 2 is that psychosocial healthcare providers are not included, but replaced by MIX to test the hypotheses.
Furthermore, measurement of religion and spirituality will be done with questionnaires that are less susceptible to ceiling effects than the HELIUS study.
Participants
The study was conducted among 46 participants. They were recruited through
churches, community centers and snowball sampling in The Hague and Amsterdam. Of these participants, five indicated that they were not Ghanaian and were therefore excluded. This left us with 41 participants (24 male, 16 female, 1 unknown, mean age = 44.29, SD = 15.76), who are comparable in age as the participants in Study 2.
Materials
Three vignettes were written to resemble a medical doctor (DR), traditional healer (TH) and medical doctor with spiritual training (MIX). The vignettes were based on
interviews conducted with healers in Ae-Ngibise et al. (2010), so the reader would read actual citations from actual traditional healers. See Appendix C for the vignettes. The vignettes had pictures and names of Ghanaian men for the sake of illustration and immersion. This
information was counterbalanced to avoid a confounding effect, which resulted in three versions of questionnaires. The participants read all three vignettes and answered the questions, which resulted in a within-design with three conditions (DR, TH an MIX).
All questions in this study had a 5-point Likert scale. There were two dependent variables per vignette: “I would go to this person if I were physically ill” and “I would go to this person if I were mentally ill”.
Furthermore, questions about the vignettes were asked to measure identification with the care provider (e.g., “I like this person”), to assess whether people in their network would go to this person, to assess whether participants trusted that this person would heal them and to assess whether the participants would be satisfied if this was his healer.
After the participants read the vignettes, they were asked to fill in the following questionnaires:
To measure spirituality, all four questions of subscale “The importance of spiritual beliefs in life” from the Spirituality Questionnaire (Parsian & Dunning, 2009) were used. One example is “My spirituality is part of my whole approach to life.”. The reliability of the subscale in the current study was .89.
To measure religiosity, a modified version of the Intrinsic/Extrinsic Measurement was used (Gorsuch & McPherson, 1989). Of the 14 questions, two questions with the lowest factor loadings were omitted (original items 5 and 14), which resulted in 12 questions from this questionnaire. The reliability of the subscale Intrinsic was .65, whereas the reliability of the subscale Extrinsic was .60.
The Abbreviated Multidimensional Acculturation Scale (AMAS, Zea, Asner-Self, Birman & Buki, 2003) was used to measure acculturation with eight questions. One example is: “I am proud to be Dutch”. The reliability of the Ghanaian subscale was .75, the reliability of the Dutch subscale was .92.
Lastly, age, gender, education level and ethnic background were asked.
Procedure
Participants were asked to fill out the questionnaire on paper, which took between 15-20 minutes. After they filled out the form, the participants were debriefed, thanked and received €2 as a reward.
Results
All analyses were conducted using SPSS (version 23) First, the effect of the vignettes were checked. A repeated measures MANOVA was conducted to investigate the effect of vignettes on the likelihood of visiting the healer in cases of a physical illness and in cases of a mental illness. Due to missing data, this analysis was conducted with 37 participants. The
multivariate test showed that there was a main effect of vignette on the intention to visit, F(2,35) = 20.241, p < .001, η² = .536 (to be interpreted below). Furthermore, there was a strong trend indicating an effect of type of illness on the intention to visit, F(1,36) = 3.787, p = .059, η² = .095, which indicates that Ghanaians have a stronger intention to visit a
healthcare provider when having a physical illness (M = 3.34, SE = .12) rather than a mental illness (M = 3.18, SE = .12). There was also an interaction effect between vignette and type of illness, F(2,35) = 5.271, p = .010, η² = .231. This indicates that the intention to visit the healer described in the vignette depends on whether the illness is physical or mental. Separate Bonferroni-corrected ANOVAs were conducted to investigate how the vignettes differed from each other on intention to visit. The ANOVAs showed that participants had a lower intention to visit TH (M = 2.43, SE = .19) than DR (M = 3.77, SE = .12), p < .001. Participant had also a weaker intention to visit TH than MIX (M = 3.58, SE = .20), p < .001. There was no difference between DR and MIX. These results show that Ghanaians generally have a stronger intention to visit DR or MIX than TH. As for the interaction, participants had a stronger intention to visit DR and MIX than TH in both physical (both p < .001) as well as mental (both p = .001) illnesses (See Table 6 for means). There is no difference between DR and MIX. Lastly, participants had a stronger intention to visit DR for physical illnesses than for mental illnesses, p = .003.
Table 6
Means and Standard Errors of Intentions to Visit Healers per Type of Issue M (SE) DR Physical Mental 4.03 (.13) 3.51 (.17) TH Physical Mental 2.35 (.21) 2.51 (.20) MIX Physical Mental 3.65 (.21) 3.51 (.19)
Several regression analyses were conducted to investigate whether religiosity and spirituality predict intentions to visit the healers. Identification was also added in the model as a new factor. This factor was derived from three questions assessing whether the
participant liked the healer, saw similarities between themselves and the healer and felt comfortable with the healer. A principal component analysis was conducted first to see whether the items would load on the same factor. Separate analyses were conducted for the three vignettes. The Kaiser-Meyer-Olkin measure verified the sampling adequacy, KMO = .64 for TH, .61 for DR and .69 for MIX (“mediocre” according to Field, 2009). Barlett’s Test of Sphericity was significant in all three cases, p <.001, indicating that correlations between items were large enough for analysis. Only one component was extracted in all analyses, explaining 70.43% of the variance for TH, 70.01% for DR and 77.42% for MIX. Regression analyses were conducted with the newly extracted factor as a measure of identification. Lastly, acculturation and questions about trusting the healer and whether people in participant’s network would go to this person were added to the regression models
to investigate whether it would predict intention to visit the healers. The correlations of these regression analyses can be found in Appendix D.
Multiple regression analysis was used to test whether the factors mentioned above predicted intention of visiting TH, DR and MIX for physical and mental illnesses. Therefore, six analyses were conducted. Each analysis was conducted the following way: first the demographic information was entered in the model (education, age and gender). In the second step, the main variables were entered into the model (identification, extrinsic and intrinsic religiosity and spirituality). In the final step, acculturation and questions regarding network and trust were included.
For intentions to visit TH in physical illnesses, the second model with seven predictors significantly predicted behavioral intention (see Table 7).
Table 7
Beta Values with their Standard Errors and Standardized Betas of the Regression Model Predicting Intention to Visit TH for Physical Illnesses
B (SE) β Step 1 Constant 4.30 (1.40) Age -.02 (.15) -.25 Gender -.35 (.46) -.13 Education -.14 (.21) -.12 Step 2 Constant 1.08 (2.45) Age -.03 (.01) -.30 Gender .01 (.41) .00
Education .06 (.20) .05 Intrinsic religiosity .03 (.06) .10 Extrinsic religiosity .14 (.08) .30
Spirituality -.07 (.07) -.18
Identification* .62 (.22) .47
R² = .07 for step 1, ∆R² = .33 for step 2 (p =.017), Model F(7,27) = 2.32, p = .039 * p < 0.05 ** p < .001
This model indicates that identification significantly predicts intention of visiting TH. Since identification is a significant predictor and none of the religiosity or spiritual measures were predictors, it is possible that the relationship of religiosity or spirituality and intention to visit is mediated through identification. A mediation analysis using the SPSS add-on PROCESS (Hayes, 2012) was conducted to test this hypothesis. A simple regression analysis showed that only extrinsic religiosity was a significant predictor of behavioral intention towards TH, therefore the mediation analysis was continued with this variable. In step 1 of the mediation model, the regression of extrinsic religiosity on intention to visit TH was significant, b = .18, t(37) = 2.57, p = .015. Step 2 showed that the regression of extrinsic religiosity on
identification was also significant, b = .15, t(37) = 2.82, p = .008. Step 3 of the mediation process showed that the regression of identification, the mediator, on intention of visiting TH while controlling for extrinsic religiosity was also significant, b = .50, t(36) = 2.43, p = .02. Step 4 indicated that while controlling for the mediator, extrinsic religiosity was not a
significant predictor of intention to visit TH, b = .10, t(37) = 1.45, p = .155. A Sobel test was conducted and did not find a full mediation in the model, z = 1.78, p =.0749. The hypothesis that identification is mediating the role of religiosity on the intention to visit TH is only partially supported.
For intention of visiting TH in mental illnesses, no model explained a significant amount of the variance.
For intention of visiting DR in physical illnesses, also no model explained a significant amount of the variance
For intention of visiting DR in mental illnesses, the third and full model explained a significant amount of the variance (see Table 8). However, none of the predictors
significantly predicted behavioral intention individually.
Table 8
Beta Values with their Standard Errors and Standardized Betas of the Regression Model Predicting Intention to Visit DR for Mental Illnesses
B (SE) β Step 1 Constant 2.50 (1.18) Age .00 (.01) .05 Gender .10 (.39) .05 Education .17 (.18) .18 Step 2 Constant .11 (2.09) Age .00 (.01) -.00 Gender .27 (.34) .12 Education .19 (.16) .20 Intrinsic religiosity .04 (.05) .13 Extrinsic religiosity .16 (.07) .42 Spirituality -.09 (.06) -.31
Identification .50 (.18) .47 Step 3 Constant -2.03 (2.54) Age .00 (.12) -.07 Gender .41 (.37) .19 Education .12 (.15) .12 Intrinsic religiosity .06 (.05) .05 Extrinsic religiosity .11 (.08) .08 Spirituality -.10 (.06) .06 Identification .39 (.25) .25 Dutch identity -.05 (.03) -.24 Ghanaian identity .15 (.08) .08 Network .09 (.37) .37 Trust -.00 (.22) .22
R² = .031 for step 1, ∆R² = .34 for step 2 (p =.017), ∆R² = .17 for step 3 (p =.11), Model F(11,23) = 2.471, p = .032
For intention of visiting MIX in physical illnesses, both the second and third model explained significant amounts of the variance. The second model seems to be optimal, since ∆R² from model 2 to model 3 is a non-significant change. As is evident from Table 9, identification is a significant predictor of the intention to visit MIX. However, none of the religious or spiritual variables directly predict intention to visit MIX, therefore a mediation is ruled out. Gender is also a significant predictor of intention to visit MIX, women seem to have a greater intention to visit MIX than men do for physical illnesses.
Table 9
Beta Values with their Standard Errors and Standardized Betas of the Regression Model Predicting Intention to Visit MIX for Physical Illnesses
B (SE) β Step 1 Constant 4,63 (1.44) Age -,01 (.15) -,15 Gender -.29 (.47) -,11 Education -,03 (.22) -,02 Step 2 Constant .44 (1.54) Age .01 (.01) .16 Gender* .58 (.27) .22 Education .37 (.13) .31 Intrinsic religiosity -.024 (.04) -.07 Spirituality .01 (.04) .02 Identification** 1.27 (.15) .96 Extrinsic religiosity .06 (.05) .12
R² = .03 for step 1, ∆R² = .73 for step 2 (p < .001), Model F(6,28) = 13.917, p < .001 * p < 0.05 ** p < .001
Lastly, for behavioral intentions towards MIX in mental illnesses, the second and third model explained significant amounts of the variance. The second model seems to be optimal, since ∆R² is a non-significant change. As evident from Table 9, identification is again a significant predictor of intention to visit MIX. None of the religious or spiritual predictors predict behavioral intention individually, therefore mediation is again ruled out.
Table 10
Beta Values with their Standard Errors and Standardized Betas of the Regression Model Predicting Intention to Visit MIX for Mental Illnesses
B (SE) β Step 1 Constant 5.14 (1.27) Age -.01 (.01) -.16 Gender -.40 (.41) -.17 Education -.155 (.19) -.15 Step 2 Constant 1.86 (1.34) Age .01 (.01) .19 Gender .34 (.24) .14 Education .15 (.11) .14 Intrinsic religiosity -.00 (.03) -.01 Spirituality .02 (.04) .07 Identification** 1.12 (.13) .96 Extrinsic religiosity -0.02 (.04) -.04
R² = .06 for step 1, ∆R² = .72 for step 2 (p < .001), Model F(7,27) = 13.038, p < .001 * p < 0.05 ** p < .001
Discussion
Ghanaians seem to prefer doctors and doctors with spiritual training over traditional healers, thus hypothesis 1 is not supported. It seems that religion and spirituality play a
smaller role than expected, an idea which is confirmed by testing hypothesis 2. Religion and spirituality do not predict intention to visit any of the healers, which does not support
hypothesis 2. Identification with traditional healers (only physical illnesses) and doctors with spiritual training predicts intention to visit them. However, this is not a mediator, which results in only a partial support of hypothesis 3. Even though the doctor is a popular option for both physical and mental illnesses, it is difficult to predict using specific predictors. Perhaps in that case, people just rely on the assumption that a purely medical doctor can handle most issues. For a doctor that also uses spiritual healing or praying, identification seems to be important. Religion or spirituality does not seem to be important in this
identification. It is possible that some Ghanaians appreciated the option of dealing with issues spiritually, without it being a leading factor.
Acculturation does not seem to predict intention to visit any healer. Lastly, gender is interestingly a significant predictor for intention to visit doctors with spiritual training in physical illnesses. Previous studies have shown that women utilize healthcare more often, however in the current study the intention to visit a healthcare provider was only higher for women in doctors with spiritual training. In all, this study is in line with Study 1 and 2, which showed no predictive effect of religion on choices in healthcare. However, identification has a predictive effect only for intentions to visit doctors with spiritual training.
General Discussion
The purpose of the presented studies was to investigate how religion, spirituality and
identification is related to healthcare choices that Ghanaians in the Netherlands make. Three hypotheses were formulated and will be discussed one by one.
Exploratory findings
First, interviews were conducted with five Ghanaians to explore the views of the Ghanaian community in the Netherlands about both traditional and medical healthcare. The
most profound finding related to religion was that they saw it as a separate, complementary addition to the treatment process. They are used to praying for health and do so in addition to medical healthcare. This does seem to support the notion of coexisting explanations, as both biological as supernatural means are used to treat illnesses. It is interesting that the
participants did not report supernatural causes of illnesses. It might be possible that it was because of the setting, where the participant was asked to be interviewed about healthcare. The first thing that comes to mind would possibly be medical healthcare, which results in a certain mindset and influences answers. Another possibility is that the small sample size is a reason no supernatural or theistic causes were reported to result into illnesses, as these are reported less often than biological causes even in religious populations (Busch et al., 2017). Furthermore, traditional healing in terms of herbal treatment does seem to be present in mostly elderly Ghanaians who are convinced it works for them, the younger generation do not longer believe in the healing powers of herbal medicine. Problems with integration does not seem to translate to problems in healthcare, as people can often find a translator or someone who can help then in the healthcare. Participants also indicated being satisfied with the healthcare system in the Netherlands. They did not indicate a lack of attention to their spirituality, as people in Ghana often do (Koenig, 2008, as cited in Osafo, 2016).
A second study investigated whether religion or acculturation could predict usage of medical, traditional and psychological healthcare in the past twelve months by using an extensive dataset. The results indicated that religion did not predict healthcare utilization. It is likely that the ceiling effect in religion influenced the results, since 88.9% of Ghanaians indicated being religious and 75.6% indicated going to a house of worship at least once a week. With this homogenous data, it is difficult to find any differences between religious and less religious people, if there is a true difference. Physical and mental wellbeing did
