Barrett-Lennard Relationship Inventory (BLRI)
by ELIZABETH FREIRE & SOTI GRAFANAK
In 1956, Barrett-Lennard was a graduate student at the Counseling Center of the University of Chicago looking for a topic for his doctoral thesis, when Rogers first circulated his theoretical formulation of the relationship conditions (one year before its publication). For his doctoral research, Barrett-Lennard decided to test Rogers‘ theory with actual clients in therapy (Barrett-Lennard, 1959). However, there were yet no measures of the therapist-to-client relationship conditions and then Barrett-Lennard had to ‘invent them from the ground up’ (Barrett-Lennard, 2002, p. 65). Barrett-Lennard reasoned that the relationship ‘as experienced by the client would be most crucially related to the outcome of therapy’ (Barrett-Lennard, 2002, p. 67).Consequently, he decided to focus his instrument on the client’s perceptions of the therapist’s attitudes in the relationship, supplemented by the therapist’s views of his/her own responses.
Description of the instrument
The BLRI comprises four subscales: ‘Empathic Understanding共情式理解’, ‘Level of Regard关注程度’, ‘Unconditionality无条件性’, and ‘Congruence一致性’. Barrett-Lennard (1962) considered that the concept of Unconditional Positive Regard (UPR) could not be treated as a unitary dimension or single variable, and therefore he separated UPR into two distinct variables: ‘Level of Regard’ and ‘Unconditionality’. In the initial version of the instrument, Barrett-Lennard (1962) had included a fifth variable called ‘Willingness to be known’ but the results for this variable were ambiguous and he decided to drop it from later versions of the inventory. However, some elements of this scale were absorbed into the Congruence dimension (Barrett-Lennard, 1978, 1986).
The BLRI is structured as a self-report questionnaire, with a six-point bipolar rating scale ranging from -3 (‘NO, I strongly feel that it is not true’) to +3 (‘YES , I strongly feel that it is true’). The 64-item BLRI (Barrett-Lennard, 1978), the version most widely used today (Barrett-Lennard, 1998; 2003), contains 16 items (8 positively worded and 8 negatively worded) for each of the four sub-scales. Examples of items from the 64-item client form (Other-to-Self, or OS) are presented in the table below.
Clients are asked to mentally insert the name of the therapist in the underlined space in each item.
SUBSCALE
ITEMS
37. Level of Regard (+)
______ is friendly and warm toward me.
33. Level of Regard (-)
______ just tolerates me.
30. Empathic Understanding (+)
_____ realises what I mean even when I havedifficulty in saying it.
58. Empathic Understanding (-)
______’s response to me is usually so fixed and automatic that I don’t get through to him/her.
51. Unconditionality (+)
Whether thoughts and feelings I express are ‘good’ or ‘bad’ makes no difference to ______’s feeling toward me.
11. Unconditionality (-)
Depending on the way I am, ______ has a better (or worse) opinion of me sometimes than at other times.
12. Congruence (+)
I feel that ______ is real and genuine with me.
52. Congruence (-)
There are times when I feel that ______’s outward response to me is quite different from the way he/she feels underneath.
The items in the therapist’s form (‘Myself-to-the-Other’, or MO) are worded in the first person for therapists to describe their response to their clients. These items are equivalent to the items in the client’s form (Barrett-Lennard, 1986). However, this equivalence is not exact because that would make the items sound ‘unnatural’ (Barrett-Lennard, 2002, p.71). The following examples (see over) of the therapist’s form (MO) correspond to like-numbered items in the client’s form (OS) listed above.
Revisions
The first version (1962) of the BLRI consisted of 85 items, but since then the instrument has undergone a number of modifications, which have resulted in a considerable reduction in the number of items. These modifications have primarily been directed toward enhancing the wording of the items and reducing response bias by balancing positively and negatively stated items. However, the essential structure and rationale of the various versions are identical to the original one (Barrett-Lennard, 2002; 2003).
1962年的第一版BLRI由85个题目构成。
MEASURING THE RELATIONSHIP CONDITIONS IN PCEP
SUBSCALE
37. Level of Regard (+)
I feel friendly and warm toward ______ .
33. Level of Regard (-)
I put up with ______ .
30. Empathic Understanding (+)
I can tell what ______ means, even when he/she has difficulty in saying it.
58. Empathic Understanding (-)
I often respond to ______ rather automatically,without taking in what he/she is experiencing.
51. Unconditionality (+)
Whether ______ is expressing ‘good’ thoughts and feelings, or ‘bad’ ones, does not affect the way I feel toward him/her.
11. Unconditionality (-)
Depending on ______’s actions, I have a better opinion of him/her sometimes than I do at other times.
12. Congruence (+)
I feel that I am genuinely myself with ______.
52. Congruence (-)
There are times when my outward response to ______ is quite different from the way I feel underneath.
The only substantial alteration of the BLRI has been the 40-item version, which has 10 items for each subscale (Barrett-Lennard, 2002). In this version, Barrett-Lennard not only dropped some items, but he also merged some others and reversed the positive/negative wording of a few items. These modifications were based more on ‘experience and judgement’ than on psychometric analysis of the items (Barrett-Lennard, 2002, p. 73).
Moreover, many distinct adaptations of the main 64-item and 40-item forms have been developed for particular uses or for specific populations. There are BLRI forms developed for students/teachers, children, groups, dyads, ‘relational life space’, supervisory relationships, nurse/patient, and doctor/patient relationships. Other further developments include an observer form (O-64) and a form for group members outside of therapy (OS-G-64) (Barrett-Lennard, 1984, 1998, 2002, 2003).
Other researchers have also added items to the original BLRI for the purposes of their own investigation. For instance, Lietaer (1976) added items related to ‘directivity’ in his Dutch-language translation, and Cramer (1986a) added an ‘advice-given’ scale to the BLRI in his study.
Validity
The initial items in the BLRI were derived from Rogers’ (1957) paper and from the Relationship Q-Sort (Bown, 1954). The content of these items were revised following discussions with the staff members at the University of Chicago Counseling Center. According to Barrett-Lennard (1962), ‘the preparation of items involved constant interaction between theory and operational expression and resulted in a continuous growth and progressive refinement of meaning relating to each concept’ (p. 6). The construct validity of the BLRI is also supported by the formal content-validation procedure carried out to eliminate non-differential items. Five qualified judges (Rogers ‘might have been’ one of them1) analyzed and checked carefully each item in order to eliminate items that did not express the variable they were designed to represent (Barrett-Lennard, 1978). The subscales were derived using a combination of item analysis and rational-theoretical considerations (Barrett-Lennard, 1959). Moreover, according to Barrett-Lennard, the considerable range of independent studies that have demonstrated an association between the BLRl and therapy outcome provides substantial evidence of ‘predictive construct validity’ (Barrett-Lennard, 1998, 2003)
Reliability
In an extensive review of the evidence on the BLRI, Gurman (1977) reported internal reliability data from 14 studies (five in actual therapy settings, four in therapy analogue settings, and five studies on other type of relationships, e.g., teachers, parents, and friends) with differing versions of the instrument. The mean internal consistency reliabilities across these 14 studies (four of them used alpha coefficients, the others used split-half reliabilities) were .91 for Level of Regard, .88 for Congruence, .84 for Empathy, .74 for Unconditionality and .91 for the total score. These results indicate that the 85-item and 64-item forms of the BLRI have high internal reliability.
There are no psychometric data reported on the 40-item version, although the reliability is expected to be a little lower, given the reduced number of items. Thus, Barrett-Lennard recommends the use of the 64-item version ‘where length is not a problem’ (Barrett-Lennard, 2002, p. 74).
Intercorrelation of the BLRI subscales
Gurman (1977) reviewed 16 studies that reported intercorrelations among the BLRI subscales and concluded that (a) Empathy, Level of Regard, and Congruence present a moderate positive correlation, i.e., these dimensions ‘appear to be relatively dependent’; (b) Unconditionality bears a very low (and in one case negative) correlation with the other dimensions (i.e., it is ‘quite independent’); and (c) Empathy, Level of Regard, and Congruence are all either moderately or highly correlated to the total score (p. 510).
Factor Analysis
Gurman (1977), after reviewing three studies that factor analysed the BLRI using item inter-correlation, concluded that the BLRI is ‘tapping dimensions that are quite consistent with Barrett-Lennard’s original work on the inventory’ (p. 513). However, he pointed out that more factor-analytic work on the BLRI on actual therapy settings should be undertaken. Almost ten years later, Cramer (1986b) factor analysed the original version of the BLRI and found that the first four factors accounted for 49.5% of the variance and reflected the four subscales postulated by the instrument. However, according to Cramer, half of the items did not “clearly distinguish the four factors”, and he concluded that “further refinement of this questionnaire was necessary to improve its factorial validity” (p. 126).
Further comments
The BLRI has been the most extensively used measure in PCE psychotherapy research. It has been considered the most suitable instrument to test Rogers’ theory of the relationship conditions since it taps into the client’s perceptions of the therapeutic relationship (Asay & Lambert, 2001; Gurman, 1977; Lockhart, 1984; Watson & Prosser, 2002). The BLRI has gained wide reputation and has been translated in many languages, including Arabic, Dutch, French, German, Greek, Italian, Japanese, Korean, Polish, Portuguese, Slovak, Spanish, and Swedish (Barrett-Lennard, 2002).
The usage of the BLRI has expanded beyond the psychotherapeutic context to wider applications in other human service contexts and significant personal life relationships (e.g. family, friendship, work, and classroom relationships). ). An important strength of the BLRI is its extensive use in clinical settings and its validation primarily in actual counselling interactions, rather than analogue settings. However, the use of different forms, modifications of content and response format, and the use of isolated sub-scales (usually empathy) rather than the whole inventory by various researchers have posed significant challenges to its further empirical validation and systematic psychometric assessment (Ponterotto, & Furlong, 1985).
－－－－－－－－－－－－－
节选自：MEASURING THE RELATIONSHIPCONDITIONS IN PERSON-CENTRED ANDEXPERIENTIAL PSYCHOTHERAPIES: PAST, PRESENT, AND FUTURE
http://vdisk.weibo.com/s/ud51QbwfhrvR
www.psychspace.com心理学空间网
