186 CHAPTER 5
(increased number of sessions) (Otto, 2002; Tang & DeRubeis, 1999). Alterna-
tively, both dose and response could be affected by some other factor. For example,
perhaps people who fare better as a result of treatment have more social support
available in their lives, and these supportive individuals encourage them to continue
treatment. As noted earlier in this chapter, correlation does not imply causation.Challenges in Researching Social Factors
Research on treatment may also investigate the possible benefi ts of targeting social
factors (for example, by matching patient and therapist by gender or ethnicity) or
the ways in which treatment may be affected by the larger social context (including
shared beliefs about a treatment’s benefi cial effects).Matching Patient and Therapist by Gender and Ethnicity
All types of psychotherapy involve a relationship between two (or more) people. Ther-
apists and patients may be similar with regard to racial or ethnic background and
gender, or they may be different. Does such a difference matter? Research suggests not:
Differences between patient and therapist in ethnicity, gender, and age do not systemati-
cally alter therapy’s benefi cial effect (Beutler, Machado, & Neufeldt, 1994; Fiorentine &
Hillhouse, 1999; Garfi eld, 1994; Lam & Sue, 2001; Maramba & Hall, 2002). How-
ever, with regard to gender, one study found that women and men are both less likely
to drop out of treatment if they have a female therapist, although the study did not
address the reason for this gender preference (Flaherty & Adams, 1998).
Nevertheless, some people prefer a therapist with a similar ethnic or racial
background to their own. For those with a strong preference, such as some Asian
Americans, matching the ethnicity of the patient and therapist may lead to better
outcomes (Sue, Zane, & Young, 1994), and it can result in lower dropout rates
among non-Whites (Flaherty & Adams, 1998; Flaskerud & Liu, 1991; Fiorentine &
Hillhouse, 1999; Sue, Kuraski, & Srinivasan, 1999). However, not all studies have
reported this result (Arcia, Sanchez-LaCay, & Fernandez, 2002).
Research on psychotherapy across different ethnic groups is complex, in part be-
cause of possible confounding factors. For instance, ethnicity is often associated with
other demographic variables, such as socioeconomic status. Moreover, a patient’s
preference for a therapist from the same ethnic group may suggest that it’s important
for the patient and therapist to share values, not the same ethnic background per se
(Ito & Maramba, 2002; Karlsson, 2005; Wong et al., 2003). In addition, research
suggests that the more culturally assimilated a patient from a nonmajority ethnic
group is, the less patient-therapist matching matters (Alvidrez, Azocar, & Miranda,
1996; Lamb & Jones, 1998). Finally, research on matching by ethnicity usually in-
volves broad categories, such as patient and therapist who are both Asian American.
However, when patients prefer a therapist from their own ethnic group, matching
them with a therapist from a broadly similar group may not suffi ce. A Korean Amer-
ican patient, for instance, may prefer a Korean American therapist, but if such a
therapist is not available, that patient may not prefer a Chinese American therapist
over a therapist of any other background (Karlsson, 2005).Culturally Sanctioned Placebo Effects
As we saw in Chapter 1, throughout time and across cultures, people in the role of
healer have used different methods to treat abnormality. For some of the problems
treated by healers—in the past and present—the placebo effect may be a key ele-
ment of successful treatment. For instance, among the Shona people of Africa, those
who have psychological problems often visit a n’anga (healer), who may recommend
herbal remedies or steam baths or may throw bones to determine the source of a
person’s “bewitchment.” Once the source is determined, the ill person’s family is told
how to mend community tensions that may have been caused by a family member’s
transgressing in some way (Linde, 2002). Westerners might consider the n’anga’s
treatments, if effective, to work because of the placebo effect and the common factors
that arise from any treatment (even a placebo): hope, support, and a framework for
understanding the problem and its resolution.Spencer Grant/Photo Edit
Some patients prefer a therapist who shares
their ethnic background, and this common
background may make a patient less likely to
drop out of treatment. However, such matching
does not appear to produce better treatment
outcomes (Beutler, Machado, & Neufeldt, 1994;
Garfi eld, 1994; Lam & Sue, 2001).