Journal of Marital and Family Therapy

doi: 10.1111/j.1752-0606.2011.00258.x

ASSESSING FAMILY CAREGIVING: A COMPARISON

OF THREE RETROSPECTIVE PARENTIFICATION
MEASURES
Lisa M. Hooper
University of Alabama

Kirsten Doehler
Elon University

Despite the frequent use of the Parentification Questionnaire (PQ) and the Parentification
Scale (PS) in research studies, scant attention has been directed toward the psychometric
properties of the scores derived from these measures. Moreover, given the importance of
parentification as a clinical topic for treatment, it is surprising that more attention has not
been paid to the accuracy and psychometric properties of parentification instruments. The
current study was the first to compare the psychometric properties of these two commonly
used parentification instruments (i.e., PQ and PS). This study also examined another
parentification instrument, newly developed: the Parentification Inventory (PI). More
specifically, the psychometric properties of the PI were compared comprehensively with
those of the PQ and PS in a sample of college students (N = 787). Scores derived from
the three measures were found to be comparable, valid, and reliable. Implications for
couple and family system practice, and directions for future research are proffered.

Parentification is ‘‘a disturbance in generational boundaries’’ (Hooper, 2007a; p. 332) that

can be evidenced by a reversal of roles within the family system (McMahon & Luthar, 2007;
Wells & Jones, 1998, 2000). In its simplest form, parentification occurs when a child assumes
adult-like roles and responsibilities (Boszormenyi-Nagy & Spark, 1973; Champion, 2005; Locke
& Newcomb, 2004). Typically, parentified children suppress and defer their own needs so that
they can meet the needs of another family member (e.g., parent, caregiver, or sibling; Carroll &
Robinson, 2000; Chase, Deming, & Wells, 1998; Fagan, 2003; Hooper, 2007a, 2011; Kelley
et al., 2007; Macfie, McElwain, Houts, & Cox, 2005b; McMahon & Luthar, 2007). Although
people who are parentified as children are not fated to experience negative outcomes in adult-
hood, the aftereffects and pernicious outcomes often seen in adulthood have been studied
empirically and reported extensively in the clinical and research literature base (Chase, 1999;
Hooper, Marotta, & Lanthier, 2008). Indeed, retrospective, self-reported parentification in
childhood is associated with deleterious outcomes in overall health and wellness across the life
span (see Chase, 1999).
Given the importance of parentification as a clinical topic for treatment, it is surprising
that more attention has not been paid to the accuracy and psychometric properties of parentifi-
cation instruments. The purpose of this study was to examine and compare the psychometric
properties of two commonly used measures of retrospective parentification—Parentification
Questionnaire (PQ; Jurkovic & Thirkield, 1998) and Parentification Scale (PS; Mika, Bergner,
& Baum, 1987)—with those of one new measure: Parentification Inventory (PI; Hooper, 2009).
Although the research suggests indirectly that scores derived from the PQ and PS yield satisfac-
tory psychometric properties, relatively few studies have purposefully and directly examined
their psychometric properties, and no studies have directly compared the psychometric proper-

Lisa M. Hooper, PhD, Department of Educational Studies in Psychology, Research Methodology, and
Counseling, University of Alabama; Kirsten Doehler, PhD, Department of Mathematics and Statistics, Elon
University.
Address correspondence to Lisa M. Hooper, Department of Educational Studies in Psychology, Research
Methodology, and Counseling, University of Alabama, Box 870231, Tuscaloosa, Alabama 35487; E-mail:
lhooper@bamaed.ua.edu

JOURNAL OF MARITAL AND FAMILY THERAPY 1

ties of the two traditional measures. Moreover, the scant attention paid to the psychometric
properties of these instruments may explain why they are used in the research base primarily
and less in clinical settings. The current study allowed for the comparison of incremental valid-
ity (Haynes & Lench, 2003) among all three retrospective, self-report parentification measures.
Additionally, the results of the study may engender confidence in the instruments among the
clinical community so that they may be used in the context of research and practice alike.
Therefore, this study was designed to fill these three important gaps: (a) to examine pur-
posefully the psychometric properties of two commonly used instruments in the research litera-
ture base that measure parentification; (b) to compare a new measure—one that attempts to
combine elements from existing measures into one comprehensive measure—with the other
measures; and (c) to report on the relations between retrospective childhood parentification and
adult functioning—both psychological health and physical health—which may engender confi-
dence in their use among therapists in the clinical and practice community.

BACKGROUND

Previous research using the PQ (Jurkovic & Thirkield, 1998) and PS (Mika et al., 1987)
has suggested that the consequences of parentification fall along a continuum from mild to
great depending on the demands placed on the child, the length of time in which the responsi-
bilities are carried out, the extent to which the parent-like roles and responsibilities are valued,
to whom the parentification roles and responsibilities are directed (i.e., parent, sibling, grand-
parent and so forth), and the personality characteristics that the child possesses. The next sec-
tion describes the accumulated indirect evidence of validity and reliability of the PQ and PS
and empirical evidence that has been informed by these measures.
According to Aldridge (2006), Hooper (2011), and Mika et al. (1987), the caretaking role
can become detrimental both in childhood and throughout the life span when it becomes a
long-term role that requires more of the child, physically and emotionally, than the child can
give because of age and maturity levels. Godsall, Jurkovic, Emshoff, Anderson, and Stanwyck
(2004), and Mika et al. also asserted that the caretaking role can become harmful to a child if
the child’s needs are neglected because of the responsibilities placed upon him or her, if the par-
ent takes on a child-like role in relation to the child and if the child is not recognized for fulfill-
ing the role or is punished for performing its duties. Because of the additional concerns and
responsibilities of caretaking, the parentified child may not learn ‘‘developmentally appropriate
tasks that build self-concept,’’ which can negatively affect a child (Godsall et al., 2004, p. 799)
and the adult that he or she later becomes (Hooper, DeCoster, White, & Voltz, 2011).
An important point highlighted by child development and family system scholars is that
childhood relationships and attachments can greatly influence relational patterns during adult-
hood (Byng-Hall, 2008a; Hooper, 2003). Parentification has been associated with higher levels
of difficulty in establishing ‘‘a separate and authentic sense of self in adult relationships and
work’’ (Chase et al., 1998, p. 106). Cree (2003) posited that some caregivers, especially younger
children, can experience difficulties because of their young caregiving roles when they transition
from adolescence to adulthood. Cree also suggested that certain problems experienced by young
caregivers increase in severity as the caregivers grow older. The problems that Cree found to
worsen with age in her young caregiving participants included the following: ‘‘sleeping difficul-
ties, eating problems, truancy, trouble with the police, substance abuse, self-harm and worries
about not having friends’’ (p. 306). Alarmingly, the risk of suicide also appears to increase as
the caregiving children grow older. In Cree’s study, four of the 61 participants between the ages
of 10 and 12 years reported having had suicidal ideation, and Cree suggested that the risk of
suicide for all young caregivers should be closely monitored.
Parentification can include duties other than emotionally focused caretaking, such as help-
ing to provide financially for one’s family. Longest and Shanahan (2007) suggested that adoles-
cents living in families with only one parent may have to work an excessive number of hours to
help support their families, which can lead to higher levels of stress and a higher likelihood of
substance use. In their study of 3,290 US adolescents, Longest and Shanahan found that ado-
lescent participation in paid work was associated with adolescent use of alcohol and sometimes

2 JOURNAL OF MARITAL AND FAMILY THERAPY

marijuana. They found that those adolescents whose parents monitored their activities were
significantly less likely to drink alcohol frequently. This buffering effect, though, was only found
present for adolescents who worked a moderate number of hours, not for those who worked an
excessive number of hours each week. In addition, parental monitoring was not found to affect
marijuana use by the working adolescents involved in Longest and Shanahan’s study. Another
important finding from that study was that younger adolescents were more likely to be protected
from substance use because of parent monitoring than were older adolescents.
Some researchers have argued that child parentification can lead to poor parenting skills
later in life (Byng-Hall, 2008b)—although Locke and Newcomb (2004), in their longitudinal
study of 318 parents (followed since childhood), found that if participants whose parents had
abused alcohol or other drugs were indeed parentified, the experience did not seem to nega-
tively affect their development of successful parenting skills. Byng-Hall (2008b) asserted that
parentified children often transmit the pattern of responsibility they faced as children to their
own children or, conversely, that they overcorrect for their childhood experiences, preventing
their children from having the opportunity to develop parenting skills. The latter pattern can
lead this next generation of children to have high demands, in turn, for their own children,
thereby transmitting parentification, although skipping one generation after the original paren-
tified child. Byng-Hall (2008b) also suggested that adults who were parentified as children often
choose spouses looking for someone to take care of them, thereby perpetuating the role of care-
taker in those adults’ lives. Alexander, Teti, and Anderson (2000) also reported that parenting
styles tend to be passed down through families as children become adults and have children of
their own.
Wells, Glickauf-Hughes, and Jones (1999) studied the relationship between parentification
and codependency in 200 university students, finding that parentified children are more likely
than nonparentified children to display codependency in adulthood. They also found that code-
pendency tends to be positively correlated with feelings of shame and inadequacy toward one-
self (see also Wells & Jones, 2000). Katz and Nelson (2007) reported that unfairness, or the
feeling that one cannot depend on parents to provide for one’s needs, is part of parentification.
Katz and Nelson, in a study of 98 North American undergraduate college students and their
parents, found that student-reported unfairness in the childhood family of origin was signifi-
cantly correlated with self-criticism later in life.

THE PRESENT STUDY

In general, self-report questionnaires are commonly used in the conduct of evidence-based

clinical practice and empirical research. A careful review of the literature (such as the foregoing
section) on parentification reveals that two commonly used instruments are the PQ (Jurkovic &
Thirkield, 1998) and the PS (Mika et al., 1987), although little focus has been explicitly placed
on their psychometric properties. Taken together, the clinical and research literature confirms
that parentification experienced in childhood is an important correlate and predictor of mental
health and functioning in adulthood, and such instruments are a vital part of understanding
this phenomenon. But the expansion of research on parentification over the past decade, the
possibility for bimodal outcomes, and changes in demography and family configurations all
warrant the development of additional parentification instruments that are informed by these
changes. Additionally, few studies have purposefully and intentionally examined the validity of
the PQ and PS. Because construct validation is an ongoing process, the current study adds to
the literature in important ways (i.e., examined scale score reliability and validity). This study
also examined another parentification instrument, newly developed: the PI.
The primary reason for the development of the PI was to provide a multi-dimensional
instrument that is undergirded by an integrative framework informed by existing measures. The
PI attempts to incorporate characteristics of the PQ (Jurkovic & Thirkield, 1998) and the PS
(Mika et al., 1987) and to add items that may be relevant to 21st-century families (e.g., the
growing phenomenon of alternate caretakers present in many families, as well as disparate cul-
tural practices) with whom family and couple therapists work. It is noteworthy that the PQ
and PS have been used primarily by researchers and scientists and less by therapists in clinical

JOURNAL OF MARITAL AND FAMILY THERAPY 3

settings. Moreover, even though indirect has accumulated for these measures (as described
above), no studies have examined their psychometric properties directly.
The purpose of this study was to compare the psychometric properties of the PQ (Jurkovic
& Thirkield, 1998), the PS (Mika et al., 1987), and the newly developed PI (Hooper, 2009).
Although empirical research has suggested that scores derived from the PQ and PS yield satis-
factory psychometric properties, no studies have compared directly the psychometric properties
of these two traditional measures. Moreover, no studies have juxtaposed the psychometric
properties of the PQ and the PS with those of the newly created instrument, the PI. This study
was designed to fill these important gaps.

AIMS AND HYPOTHESES

Three specific aims motivated this study. Aim 1 sought to examine the convergent validity
of the three assessments of parentification: PI, PQ, and PS. Aim 2 sought to examine the inter-
nal consistency (Cronbach’s alphas) of the PI, PQ, and PS. Finally, aim 3 sought to investigate
the divergent and predictive validity of the PI, PQ, and PS by relating the scores from these
measures to those from the SF-36 health outcome measure.
Three corresponding hypotheses were, therefore, examined in this study:
Hypothesis 1. The PI, PQ, and PS will be strongly positively correlated with one another.
Hypothesis 2. Scores from the PI will result in reliability coefficients that (a) are similar to
the PQ and PS and (b) are 0.70 or greater.
Hypothesis 3. The higher scores on the parentification measures will be correlated with and
predictive of scores derived from a measure of psychological and physical health (i.e., SF-36).

METHOD

Survey Procedure
Following Institutional Review Board approval, participants were recruited to take part in
a study investigating the link between childhood roles and responsibilities and adult functioning.
The survey was administered online using a web-based methodology. Specifically, participants
were sent an electronic invitation to participate in the study. The electronic invitation included a
description of the study and a direct link to the electronic survey and the informed consent form.
Four instruments and a demographic information sheet were used in the current study. On aver-
age, participants took approximately 35–40 min to complete the questionnaires. All instruments
were in English. Extra course credit was provided as an incentive for participating in the study.
Participants received reminder e-mails 2 weeks after they expressed an interest in the study.

Measures
Demographic information sheet. This instrument, created for the study, asked survey par-
ticipants to respond to background questions. Questions covered the participant’s year in
school, academic program, age, gender, and racial and ethnic background.
Parentification Inventory. The PI (Hooper, 2009) is a retrospective 22-item self-report mea-
sure that assesses caregiving roles and responsibilities usually reserved for adults but carried out
by children. The PI is also designed to measure the perceived benefits of performing caregiving
roles in one’s family of origin. Participants respond to the 22 items using a five-point Likert-type
scale, ranging from 1 (never true) to 5 (always true). The PI consists of three subscales: parent-
focused parentification, sibling-focused parentification, and perceived benefit finding of parentifi-
cation. Items associated with parent-focused parentification include, for example, ‘‘I was expected
to comfort my parents when they were sad or having emotional difficulties’’ and ‘‘My parent(s)
often shared secrets with me about other family members.’’ Items associated with sibling-focused
parentification include, for example, ‘‘I was responsible for making sure that my siblings went to
bed every night’’ and ‘‘I was the primary person who disciplined my siblings.’’ Items associated
with perceived benefit finding of parentification include ‘‘I really enjoyed my role in the family.’’
Scores can fall in the range of 1–5, with higher total and subscale scores reflecting greater
perceived levels of parentification or perceived benefits of the parentification process.

4 JOURNAL OF MARITAL AND FAMILY THERAPY

 The initial phase of the development and item generation of the PI was informed by the
survey development literature base (Worthington & Whitaker, 2006), by a comprehensive review
of family system theory, and by other instruments that have captured elements of childhood paren-
tification retrospectively (Jurkovic & Thirkield, 1998; Mika et al., 1987). In the original research,
exploratory and confirmatory factor analysis resulted in a three-factor solution for the PI items.
Internal consistency coefficients evinced in the confirmatory factor analysis ranged from 0.79 to
0.84. Further details of the factor analysis can be obtained from the first author of this paper.
Parentification Questionnaire. The PQ (Jurkovic & Thirkield, 1998) is a widely used 30-
item self-report instrument that retrospectively measures three dimensions of perceived parentif-
ication: instrumental parentification, emotional parentification, and perceived fairness of the
parentification process. Of the 30 items, 10 pertain to instrumental parentification, 10 pertain to
emotional parentification, and 10 pertain to perceived fairness. Participants rate how true the
statements are on a five-point Likert scale, where 1 is strongly disagree and 5 is strongly agree.
The overall score for parentification, therefore, falls in the range of 30–150. Subscales—
instrumental parentification, emotional parentification, and perceived fairness—can yield a
score in the range of 10–50. For both the overall score and the subscale scores, higher scores
reflect greater parentification and ⁄ or perceived fairness. Cronbach’s alphas for the PQ subscale
scores have been reported to be in the range of 0.82–0.92 (Burnett, Jones, Bliwise, & Ross,
2006; Hooper & Wallace, 2010; Jurkovic, Thirkield, & Morrell, 2001; Kelley et al., 2007).
Parentification Scale. The PS (Mika et al., 1987) is the second most widely used instru-
ment that captures parentification. Designed to assess four types of parentification (child par-
enting parent[s], child acting as spouse to parent, child parenting siblings, and child taking on
other roles generally taken by adults), this self-report assessment includes 30 items asking adults
to indicate how often they fulfilled a particular adult responsibility as a child. Items are
summed and computed to create subscale scores according to responsibilities. Cronbach’s
alphas for the PS subscale scores have been reported to be in the range of 0.76–0.86 (Fitzgerald
et al., 2008). All subscales were used in the current study.
SF-36 Health Outcome Survey. The SF-36 (Ware & Sherbourne, 1992) assesses a range of
health-related conditions and constructs (e.g., physical functioning, general mental health, social
functioning, and bodily pain). The 36-item self-report survey generates eight subscale scores
that are combined to create two aggregate summary scores of health functioning—mental
health and physical health. The two summary scores, which were used in the current study,
assess respondents’ overall level of physical functioning and overall level of mental functioning.
The psychometric properties of the SF-36 survey are reported to be excellent. This instru-
ment has a long history of being both highly reliable and valid. Cronbach’s alpha coefficient
has yielded scores on the eight subscale and two summary scores in the range of 0.82–0.92
(Ware, Kosinski, & Keller, 1994). Test–retest reliability has yielded 0.69–0.91 for the eight sub-
scale scores and 0.80–0.90 for the summary scores (Ware et al., 1994).

RESULTS

Sample Characteristics
The convenience study sample consisted of 787 student volunteers recruited from under-
graduate classes in a southeastern university. Participants’ ages ranged from 19 to 48 years
(M = 20.86, SD = 3.55). The sample primarily consisted of White Americans (82.5%,
n = 649) and females (76%, n = 598). Self-reported races of the remaining participants were
American Indian (0.05%, n = 4), Asian American (1.0%, n = 8), African American (12.5%,
n = 98), Hispanic–Latino (0.04%, n = 3), mixed race (3.0%, n = 21), and those who failed to
report a race (0.05%, n = 4).

Correlations Among Scales

The Pearson correlation r was used to measure effect size, representing the correlation
between the parentification measures. Table 1 shows the bivariate correlations between the
measures (PI, PQ, and PS). As expected, there were significant positive correlations between
the subscale scores on the PI, PQ, and PS. The magnitude of these relations varied from small

JOURNAL OF MARITAL AND FAMILY THERAPY 5

 6
Table 1
Bivariate Correlation Matrix for Parentification Measures (PI, PQ, and PS; Hypothesis 1; N = 787)

Variables 1 2 3 4 5 6 7 8 9 10

1. PI (PFP) –
2. PI (SFP) 0.421** –
3. PI (PBP) 0.264** 0.147** –
4. PQ (IM) 0.540** 0.538** 0.352** –
5. PQ (EM) 0.683** 0.310** 0.455** 0.617** –
6. PQ (PF) 0.466** 0.259** 0.688** 0.637** 0.694** –
7. PS (SR) 0.561** 0.244** 0.352** 0.440** 0.632** 0.479** –
8. PS (PR-P) 0.580** 0.248** 0.298** 0.430** 0.631** 0.449** 0.841** –
9. PS (PR-S) 0.375** 0.564** 0.260** 0.538** 0.382** 0.371** 0.521** 0.537** –
10. PS (NS) 0.394** 0.387** 0.272** 0.567** 0.409** 0.394** 0.447** 0.460** 0.503** –

Note. PI = Parentification Inventory (PFP = parent-focused parentification; SFP = sibling-focused parentification; PBP = perceived benefits of
parentification); PQ = Parentification Questionnaire (IM = instrumental parentification; EM = emotional parentification; PF = perceived fairness);

JOURNAL OF MARITAL AND FAMILY THERAPY

PS = Parentification Scale (SR = spousal role; PR-P = parental role with parent; PR-S = parental role with sibling; NS = nonspecific).
**p < .01.
to high, with associations ranging from 0.26 to 0.84. Therefore, support was found for
Hypothesis 1.

Scale Score Reliability

The reliability of all instruments’ scores was determined by calculating Cronbach’s alpha
coefficients. The subscale scores of the parentification measures were uniformly high and consis-
tent across all measures in the current study sample. Hypothesis 2 suggested that the measures’
scores would meet the conventional standard for adequate reliability (i.e., Cronbach’s alpha of
.70 or greater) in the current study sample. As seen in Table 2, the results support this hypothe-
sis: Cronbach’s alphas for the subscale scores for all three scales were reported to be in the
range of 0.76–0.91. Therefore, support was found for Hypothesis 2.

Scale Score Validity

The extent to which PI scores measure the intended construct—parentification—was also
assessed by examining the extent to which PI scores related to other constructs (i.e., psychologi-
cal and physical health outcomes) in similar and theorized ways compared with the PQ and the
PS. As shown in Table 2, the results derived from all three parentification measures are fairly
consistent and similar. In general, and as expected, the relations between all PSs and subscales
and the SF-36 subscales indicated that participants with higher levels of parentification also
reported poorer health outcomes; in other words, the higher the level of parentification, the
lower the level of positive psychological and physical health. Additionally, divergent validity
was established, in that the association between the parentification measures was greater than
the associations evinced between the parentification measures and the SF-36 subscales. Thus,
these findings—in conjunction with scale score predictability described elsewhere—provide sup-
port for Hypothesis 3.

Scale Score Predictability

Six separate regression analyses were performed to examine the ability of the parentifica-
tion measures to predict health outcomes. The SF-36 criterion scores were regressed on the PI
subscale scores, the PQ subscale scores, and the PS subscale scores. Because of the exploratory
nature of this study, all variables were entered simultaneously. The multiple correlation squared
(R2) and standardized regression coefficients were used to measure effect size.
Parentification Inventory. In two separate regression tests, the SF-36 mental health (test 1)
and physical health (test 2) outcome scores were regressed on the three PI predictor variables
(see Table 3). The first overall regression model established the existence of a significant rela-
tionship between the three predictor variables and the criterion variable of psychological
health: F(3, 787) = 59.50, p < .0001. The R value (r = .431) for this model meets the crite-
rion for a medium effect size, as defined by Cohen (1992; Cohen & Cohen, 1983). The R2
value reveals that 19% of the variance observed in the criterion variable of psychological
health was explained by the model. Examination of the t tests on each beta weight showed
that parent-focused parentification subscale scores and perceived benefits of parentification
subscale scores made a significant, unique contribution above and beyond the other variables:
b = .114, t (1, 787) = 3.12, p < .01 and b = .390, t (1, 787) = 11.65, p < .0001, respec-
tively.
In the second regression model, the data established the existence of a significant relation-
ship between the three predictor variables and the criterion variable of physical health: F(3,
787) = 32.43, p < .0001. The R value (r = .332) for this model meets the criterion for a med-
ium effect size, as defined by Cohen (1992; Cohen & Cohen, 1983). The R2 value reveals that
11% of the variance observed in the criterion variable of growth was explained by the model.
Examination of the t tests on each beta weight showed that parent-focused parentification sub-
scale scores and perceived benefits of parentification subscale scores made a significant, unique
contribution above and beyond the other variables: b = .173, t (1, 787) = 4.52, p < .0001 and
b = .231, t (1, 787) = 6.61, p < .0001, respectively.
Parentification Questionnaire. In two separate regression tests, the SF-36 mental health
and physical health outcome scores were regressed on the three PQ predictor variables (i.e.,

JOURNAL OF MARITAL AND FAMILY THERAPY 7

 8
Table 2
Descriptive Statistics and Correlation Matrix for Parentification Measures and Theorized Mental and Physical Health Correlates (Hypothesis
2; N = 787)

Variables 1 2 3 4 5 6 7 8 9 10 11 12

1. PI (PFP) –
2. PI (SFP) 0.421** –
3. PI (PBP) 0.264** 0.147** –
4. PQ (IM) 0.540** 0.538** 0.352** –
5. PQ (EM) 0.683** 0.310** 0.455** 0.617** –
6. PQ (PF) 0.466** 0.259** 0.688** 0.637** 0.694** –
7. PS (SR) 0.561** 0.244** 0.352** 0.440** 0.632** 0.479** –
8. PS (PR-P) 0.580** 0.248** 0.298** 0.430** 0.631** 0.449** 0.841** –
9. PS (PR-S) 0.375** 0.564** 0.260** 0.538** 0.382** 0.371** 0.521** 0.537** –
10. PS (NS) 0.394** 0.387** 0.272** 0.567** 0.409** 0.394** 0.447** 0.460** 0.503** –
11. SF-36 MENT 0.210** 0.091* 0.418** 0.186** 0.281** 0.399** 0.256** 0.225** 0.148** 0.163** –
12. SF-35 PHYS 0.244** 0.133** 0.280** 0.253** 0.206** 0.294** 0.288** 0.240** 0.217** 0.128** 0.447** –
M 1.90 1.545 4.10 17.81 22.57 17.87 32.67 24.83 51.38 15.04 975.67 1794.00
SD 0.585 0.560 0.840 6.59 7.15 7.56 6.40 4.73 8.41 3.52 238.77 293.76
Cronbach’s Alpha 0.85 0.82 0.76 0.83 0.85 0.90 0.88 0.81 0.91 0.83 0.80 0.90

JOURNAL OF MARITAL AND FAMILY THERAPY

Note. PI = Parentification Inventory (PFP = parent-focused parentification; SFP = sibling-focused parentification; PBP = perceived benefits of
parentification); PQ = Parentification Questionnaire (IM = instrumental parentification; EM = emotional parentification; PF = perceived fairness);
PS = Parentification Scale (SR = spousal role; PR-P = parental role with parent; PR-S = parental role with sibling; NS = nonspecific).
*p < .05. **p < .01.
 Table 3
Regression Analysis Summary for Parentification Measures (PI, PQ, and PS) and
Mental and Physical Health Measures (Hypothesis 3)

PI scale scores B SE B b

Outcome variable: SF-36 mental healtha

PI (PFP) 3.87 1.24 .114**
PI (SFP) 1.37 3.28 .015
PI (PBP) 36.94 3.16 .390***
PI scale scores B SE B b
Outcome variable: SF-36 physical healthb
PI (PFP) 7.21 1.59 .173***
PI (SFP) 2.93 4.22 .026
PI (PBP) 26.95 4.07 .231***
PQ scale scores B SE B b
Outcome variable: SF-36 mental healthc
PQ (IM) 4.68 1.61 .129**
PQ (EM) 1.68 1.59 .051
PQ (PF) 14.11 1.53 .447***
PQ scale scores B SE B b
Outcome variable: SF-36 physical healthd
PQ (IM) 5.41 2.07 .122**
PQ (EM) 1.53 2.04 .037
PQ (PF) 9.42 1.97 .243***
PS scale scores B SE B b
Outcome variable: SF-36 mental healthe
PS (SR) 8.11 2.41 .218***
PS (PR-P) .86 3.31 .017
PS (PR-S) .10 1.24 .004
PS (NS) 4.01 2.81 .059
PS scale scores B SE B b
Outcome variable: SF-36 physical healthf
PS (SR) 12.76 2.93 .278***
PS (PR-P) 2.33 4.03 .038
PS (PR-S) 3.84 1.51 .110**
PS (NS) 2.83 3.41 .034

Note. PI = Parentification Inventory (PFP = parent-focused parentification; SFP = sib-

ling-focused parentification; PBP = perceived benefits of parentification); PQ = Parentif-
ication Questionnaire (EM = emotional parentification; IM = instrumental
parentification; PF = perceived fairness); PS = Parentification Scale (SR = spousal role;
PR-P = parental role with parent; PR-S = parental role with sibling; NS = nonspecific).
a
Model 1: R2 = .185 (n = 787, p < .0001). bModel 2: R2 = .111 (n = 787, p < .0001).
c
Model 3: R2 = .168 (n = 787, p < .0001). dModel 4: R2 = .094 (n = 787, p < .0001).
e
Model 5: R2 = .068 (n = 787, p < .0001). fModel 6: R2 = .091 (n = 787, p < .0001).
*p < .05. **p < .01. ***p < .001.

subscale scores). Once again, because of the exploratory nature of this study, all variables were
entered simultaneously. As shown in Table 3, the first overall regression model (test 1) estab-
lished the existence of a significant relationship between the three predictor variables and the
criterion variable of psychological health: F(3, 787) = 52.93, p < .0001. The R value
(r = .411) for this model meets the criterion for a medium effect size, as defined by Cohen

JOURNAL OF MARITAL AND FAMILY THERAPY 9

(1992; Cohen & Cohen, 1983). The R2 value reveals that 17% of the variance observed in the
criterion variable of psychological health was explained by the model. Examination of the t
tests on each beta weight showed that instrumental parentification subscale scores and per-
ceived fairness of parentification subscale scores made a significant, unique contribution above
and beyond the other variables: b = .129, t (1, 787) = 2.90, p < .01 and b = .447, t (1,
787) = 9.17, p < .0001, respectively.
The second regression model (test 2) established the existence of a significant relationship
between the four predictor variables and the criterion variable of physical health: F(3,
787) = 27.27, p < .0001. The R value (r = .308) for this model meets the criterion for a med-
ium effect size, as defined by Cohen (1992; Cohen & Cohen, 1983). The R2 value reveals that
9% of the variance observed in the criterion variable of growth was explained by the model.
Examination of the t tests on each beta weight showed that instrumental parentification sub-
scale scores and perceived fairness of parentification subscale scores made a significant, unique
contribution above and beyond the other variables: b = .122, t (1, 787) = 2.61, p < .01 and
b = .243, t (1, 787) = 4.77, p < .0001, respectively.
Parentification Scale. In the final two regression tests, the SF-36 mental health and physi-
cal health outcome scores were regressed on the four PS predictor subscale scores. The first
overall regression model (test 1) established the existence of a significant relationship between
the four predictor variables and the criterion variable of psychological health: F(4,
787) = 14.44, p < .0001. The R value (r = .262) for this model meets the criterion for a med-
ium effect size, as defined by Cohen (1992; Cohen & Cohen, 1983). The R2 value reveals that
7% of the variance observed in the criterion variable of psychological health was explained by
the model. Examination of the t tests on each beta weight showed that child-as-parent subscale
scores made a significant, unique contribution above and beyond the other variables: b = .218,
t (1, 787) = 3.36, p < .0001.
In the second regression model (test 2), the data established the existence of a significant
relationship between the four predictor variables and the criterion variable of physical health:
F(4, 787) = 19.56, p < .0001. The R value (r = .302) for this model meets the criterion for a
medium effect size, as defined by Cohen (1992; Cohen & Cohen, 1983). The R2 value reveals
that 9% of the variance observed in the criterion variable of growth was explained by the
model. Examination of the t tests on each beta weight showed that child-as-parent subscale
scores and sibling-as-parent subscale scores made a significant, unique contribution above and
beyond the other variables: b = .278, t (1, 787) = 4.34, p < .0001 and b = .110, t (1,
787) = 2.54, p < .01, respectively.

DISCUSSION

This study compared the psychometric properties of three measures that assess childhood
parentification retrospectively. We found evidence of reliability and validity for all three mea-
sures. In addition, the results of the study suggest that scores from the newly created PI have
satisfactory reliability and validity. The main results also reveal that the PI generates findings
comparable to those of the PQ and PS. The main results of the study can be described as fol-
lows: First, all three measures of parentification are capturing a multi-dimensional construct
similarly. Second, all three measures’ scores are reliable, as measured by Cronbach’s alphas.
Third, all three measures’ scores demonstrate sound predictive, convergent, and divergent valid-
ity in theoretically expected ways.
Moreover, although we expected all three measures to reveal similar relations between
study variables, we also expected the parentification measures to reveal differences. For exam-
ple, an important finding from these results is the slight incremental validity (Haynes & Lench,
2003) evinced in the PI scores compared with the PQ and PS scores. The slight difference
between the three parentification instruments, as evidenced in these results, may point to an
advantage of the PI in comparison with the PQ and PS, although we are cautious in this preli-
minary finding. Further, because of the conditional nature of incremental validity (Haynes &
Lench, 2003), many more studies comprising diverse study samples are needed to fully clarify
this tentative assertion.

10 JOURNAL OF MARITAL AND FAMILY THERAPY

 IMPLICATIONS FOR COUPLE AND FAMILY THERAPISTS

The prevalence of parentification in childhood has implications for the individuals that
children become in adulthood (DiCaccavo, 2006). Moreover, the construct of parentification
has relevance to science and practice. Couple and family therapists as well as other mental
health care providers would benefit from assessing for parentification. DiCaccavo outlines
important issues relevant to the assessment of signs and symptoms in individuals and couples
who report a childhood history of parentification. Specifically, she reports there are three areas
that should be examined in a comprehensive assessment of a client with a possible history of
parentification: (a) the type of role assignment (i.e., instrumental and emotional parentification)
and the level of support received when the client performed the parentification role; (b) the age
of the client when the parentification process took place and the extent of the parentification-
related responsibilities in which the client engaged; and (c) ‘‘the consideration of the object of
concern and degree of internalization’’ (DiCaccavo, 2006, p. 472). Clients who have internalized
the parentified role and responsibilities may describe issues and concerns about others but not
about him or herself. Deparentification based on a comprehensive assessment may include
explorations related to issues of shame, splitting, and projective identification, for example (see
DiCaccavo, 2006).
Because parentification is a common clinical phenomenon likely to be observed by family
therapists as well as other mental health care providers (Byng-Hall, 2008b; Chase, 1999;
Hooper, 2007b), it is critical that useful measures with sound psychometric properties be avail-
able to assess for parentification. Additionally, a panoply of evidence (summarized earlier in
this article) supports the correlation between parentification and psychological health. The cur-
rent study adds to the clinical literature base by considering the link between parentification
and physical health in addition to the link between parentification and psychological health.
Thus, these findings are pertinent to both the physical and the psychological health of patients
with whom couple and family therapists work. The results also add to the current literature
base by describing the psychometric properties of three instruments that produce reliable and
valid scores. We hope that these results will engender therapists’ confidence in the measures
that capture parentification, such that therapists will use these instruments in their practice.

IMPLICATIONS FOR FUTURE RESEARCH

These findings strengthen the validity and empirical evidence that underpins the construct
of parentification. These results suggest some implications for future research. Over the past
30 years, an extensive amount of empirical research on parentification and possible correlates,
antecedents, predictors, and outcomes has been conducted. Less work has focused on the mea-
surement of parentification. There remains a lack of information on the test–retest reliability
and social desirability of all three instruments included in the current study. Future research
should include a measure that examines the extent to which respondents answer questions in a
way that they believe is favorable (e.g., Social Desirability Scale; Crowne & Marlowe, 1960).
Certainly, the stability and validity of the scores derived from these instruments could be
affected if respondents alter answers based on their desire to appear in a positive light.
Additionally, few studies have advanced ideas about how parentification measures may be
used in process and outcome research. Additional value—specifically clinical value—could be
garnered from establishing the clinical utility of the PI, PQ, and PS instruments. For example,
important clinical information could be uncovered if measures of parentification included in
randomized trials focused on the efficacy and effectiveness of couple and family system thera-
pies. With regard to process research, it could be that a history of parentification might impede
the joining process or building of rapport in therapy. Similar to the link between attachment
style and the relationship patients have with important individuals in their life (e.g., significant
romantic other, therapist, primary care physician), so too could a history of childhood parentif-
ication impinge upon the ability of patients to form therapeutic relationships with their thera-
pist and thereby have an impact therapeutic outcomes. Finally, few studies have investigated
potential mediators and moderators between childhood parentification and adult functioning.

JOURNAL OF MARITAL AND FAMILY THERAPY 11

Uncovering modifiable or alterable factors regarding outcomes associated with parentification
is also an important direction for future research.

STUDY LIMITATIONS

The single-informant, cross-sectional design of the current study yielded several limitations.
First, given that the current sample consisted of undergraduate students, the results may not
generalize to other populations. Second, the study sample was homogenous. Because reliability
and validity are context dependent (Schmidt & Hunter, 2003), it is important to continue to
accumulate evidence related to cultural validity and generalizability in studies of diverse popu-
lations. Researchers (e.g., Bravo, 2003) suggest that validity and reliability of an instrument’s
scores in one culture do not portend validity and reliability in another culture. Researchers
should examine the cross-cultural, linguistic, and translational equivalence of the PI, PQ, and
PS. Findings from such studies would inform needed changes in measures of parentification.
Further replications of the current study’s findings are needed to determine the extent to which
these findings may be applicable to and clinically meaningful for a range of populations.
The fact that all data in the study were based on self-report measures is another limitation
of the study. Future studies should employ multi-method and direct observation strategies to
further clarify the validity and reliability of the PI, PQ, and PS. Other considerations and possi-
ble limitations related to the measurement of parentification include retrospective recall. Mem-
ory or motivational factors may affect accurate reporting of retrospective parentification. For
example, the stability of reporting of adverse events, particularly when based on family of ori-
gin or lifetime experiences, may be confounded by memory loss, underreporting, overreporting,
or minimizing (Baker, 2009; Briere & Conte, 1993; Hooper & Wallace, 2010). Alternately, indi-
viduals may be unwilling to report events such as childhood neglect, adversity, or mistreatment
(Baker, 2009; Baker & Festinger, 2011; Brewin, Andrews, & Gotlib, 1993). Moreover, studies
that include assessments that are cross-situational sources will enhance our understanding of
parentification, the process, and the correlates and predictors of parentification (Kuperminc,
Jurkovic, & Casey, 2009).

CONCLUSION

Despite the frequent use of the PQ and the PS in research studies, scant attention has been
directed toward the psychometric properties of the scores derived from these measures. The cur-
rent study was the first to compare the psychometric properties of these two commonly used
parentification instruments. This study also examined another parentification instrument, newly
developed: the PI. The psychometric properties of the PI were compared comprehensively to
those of the PQ and PS. The current study adds to the literature base because, in addition to
being often studied, parentification is a ubiquitous clinical phenomenon that most family thera-
pists and other mental health care providers will face (DiCaccavo, 2006). Findings demon-
strated that parentification is relevant to adults’ psychological and physical health and warrants
the attention of practitioner–scientists in clinical settings. Therefore, continued examination of
psychometrically sound measures is needed to further our understanding of the parentification
process in both clinical and research settings, the outcomes across the life span that are associ-
ated with parentification and interventions that are culturally relevant and linguistically appro-
priate for racially and ethnically diverse populations.

REFERENCES

Aldridge, J. (2006). The experiences of children living with and caring for parents with mental illness. Child Abuse
Review, 15, 79–88. DOI: 10.1002/car.904.
Alexander, P. C., Teti, L., & Anderson, C. L. (2000). Childhood sexual abuse history and role reversal in parent-
ing. Child Abuse & Neglect, 24, 829–838.
Baker, A. J. L. (2009). Adult recall of childhood psychosocial maltreatment: Definitional strategies and chal-
lenges. Children and Youth Services Review, 31, 703–714.

12 JOURNAL OF MARITAL AND FAMILY THERAPY

Baker, A. J. L., & Festinger, T. (2011). Adult recall of childhood psychological maltreatment: A comparison of
five scales. Journal of Aggression, Maltreatment, and Trauma, 20, 70–89. DOI: 10.1080/
10926771.2011.538006.
Boszormenyi-Nagy, I., & Spark, G. (1973). Invisible loyalties: Reciprocity in intergenerational family therapy.
Hagerstown, MD: Harper & Row.
Bravo, M. (2003). Instrument development: Cultural adaptations in ethnic minority research. In G. Bernal,
J. Trimble, A. Burlew & F. Leong (Eds.), Handbook of racial ethnic minority psychology (pp. 220–232).
Newbury Park, CA: Sage.
Brewin, C., Andrews, B., & Gotlib, I. (1993). Psychopathology and early experience: A reappraisal of retrospec-
tive reports. Psychological Bulletin, 13(1), 82–98.
Briere, J., & Conte, J. (1993). Self-reported amnesia for abuse in adults molested as children. Journal of Trau-
matic Stress, 6, 21–31.
Burnett, G., Jones, R. A., Bliwise, N. G., & Ross, L. T. (2006). Family unpredictability, parental alcoholism, and
the development of parentification. The American Journal of Family Therapy, 34, 181–189. DOI: 10.1080/
01926180600550437.
Byng-Hall, J. (2008a). The crucial roles of attachment in family therapy. Journal of Family Therapy, 30, 129–146.
Byng-Hall, J. (2008b). The significance of children fulfilling parental roles: Implications for family therapy. Jour-
nal of Family Therapy, 30, 147–162.
Carroll, J. J., & Robinson, B. E. (2000). Depression and parentification among adults as related to parental
workaholism and alchoholism. The Family Journal: Counseling and Therapy for Couples and Families, 8,
360–367.
Champion, J. E. (2005). The role of caretaking in children of mothers with a history of depression (Master’s thesis).
Vanderbilt University, Nashville, TN.
Chase, N. (1999). An overview of theory, research, and societal issues. In N. Chase (Ed.), Burdened children
(pp. 3–33). New York: The Guilford.
Chase, N. D., Deming, M. P., & Wells, M. C. (1998). Parentification, parental alcoholism, and academic status
among young adults. The American Journal of Family Therapy, 26, 105–114.
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159.
Cohen, J., & Cohen, P. (1983). Applied multiple regression ⁄ correlation analysis for the behavioral sciences, 2nd
edn. Hillsdale, NJ: Erlbaum Associates.
Cree, V. E. (2003). Worries and problems of young carers: Issues for mental health. Child and Family Social
Work, 8, 301–309.
Crowne, D. P., & Marlowe, D. (1960). A new scale of social desirability independent of psychopathology. Journal
of Consulting Psychology, 24, 349–354.
DiCaccavo, A. (2006). Working with parentification: Implications for clients and counselling psychologists. Psy-
chology and Psychotherapy: Theory, Research and Practice, 79, 469–478. DOI: 10.1348/147608305X57978.
Fagan, M. A. (2003). Exploring the relationship between maternal migraine and child functioning. Headache, 43,
1042–1048.
Fitzgerald, M. M., Schneider, R. A., Salstrom, S., Zinzow, H. M., Jackson, J., & Fossel, R. V. (2008). Child sex-
ual abuse, early family risk, and childhood parentification: Pathways to current psychosocial adjustment.
Journal of Family Psychology, 22, 320–324. DOI: 10.1037/0893-3200.22.2.320.
Godsall, R. E., Jurkovic, G. J., Emshoff, J., Anderson, L., & Stanwyck, D. (2004). Why some kids do well in bad
situations: Relation of parental alcohol misuse and parentification to children’s self-concept. Substance Use
& Misuse, 39, 789–809. DOI: 10.1081/JA-120034016.
Haynes, S. N., & Lench, H. C. (2003). Incremental validity of new clinical assessment measures. Psychological
Assessment, 15, 456–466.
Hooper, L. M. (2003). Parentification, resiliency, secure adult attachment style, and differentiation of self as pre-
dictors of growth among college students. Dissertations Abstracts International, 64, 1493.
Hooper, L. M. (2007a). The application of attachment theory and family systems theory to the phenomena of
parentification. The Family Journal, 15, 217–223.
Hooper, L. M. (2007b). Expanding the discussion regarding parentification and its varied outcomes: Implications
for mental health research and practice. Journal of Mental Health Counseling, 29, 322–337.
Hooper, L. M. (2009). Parentification Inventory. Available from L. M. Hooper, Department of Educational
Studies in Psychology, Research Methodology, and Counseling, The University of Alabama, Tuscaloosa,
AL 35487.
Hooper, L. M. (2011). Parentification. In R. J. R. Levesque (Ed.), Encyclopedia of adolescence. New York, NY:
Springer.
Hooper, L. M., DeCoster, J., White, N., & Voltz, M. L. (2011). Characterizing the magnitude of the relation
between self-reported childhood parentification and adult psychopathology: A meta-analysis. Journal of
Clinical Psychology, 67, 1028–1043.

JOURNAL OF MARITAL AND FAMILY THERAPY 13

Hooper, L. M., Marotta, S. A., & Lanthier, R. P. (2008). Predictors of growth and distress following parentification
among college students. Journal of Child and Family Studies, 17, 693–705.
Hooper, L. M., & Wallace, S. (2010). Evaluating the parentification questionnaire: Psychometric properties and
psychopathology correlates. Contemporary Family Therapy: An International Journal, 32, 52–68.
Jurkovic, G. J., & Thirkield, A. (1998). Parentification questionnaire. Available from G. J. Jurkovic, Department
of Psychology, Georgia State University, University Plaza, Atlanta, GA 30303.
Jurkovic, G. J., Thirkield, A., & Morrell, R. (2001). Parentification of adult children of divorce: A multidimen-
sional analysis. Journal of Youth and Adolescence, 30, 245–257.
Katz, J., & Nelson, R. A. (2007). Family experiences and self-criticism in college students: Testing and model of
family stress, past unfairness, and self-esteem. The American Journal of Family Therapy, 35, 447–457. DOI:
10.1080/01926180601057630.
Kelley, M. L., French, A., Bountress, K., Keefe, H. A., Schroeder, V., Steer, K., et al. (2007). Parentification and
family responsibility in the family of origin of adult children of alcoholics. Addictive Behaviors, 32, 675–685.
DOI: 10.1016/j.addbeh.2006.06.010.
Kuperminc, G. P., Jurkovic, G. J., & Casey, S. (2009). Relations of filial responsibility to the personal and social
adjustment of Latino adolescents from immigrant families. Journal of Family Psychology, 23, 14–22.
Locke, T. F., & Newcomb, M. D. (2004). Child maltreatment, parent alcohol- and drug-related problems, poly-
drug problems, and parenting practices: A test of gender differences and four theoretical perspectives. Jour-
nal of Family Psychology, 18, 120–134.
Longest, K. C., & Shanahan, M. J. (2007). Adolescent work intensity and substance use: The meditational and
moderational roles of parenting. Journal of Marriage and Family, 69, 703–720.
Macfie, J., McElwain, N. L., Houts, R. M., & Cox, M. J. (2005b). Intergenerational transmission of role reversal
between parent and child: Dyadic and family systems internal working models. Attachment & Human Devel-
opment, 7, 51–65.
McMahon, T. J., & Luthar, S. S. (2007). Defining characteristics and potential consequences of caretaking bur-
den among children living in urban poverty. American Journal of Orthopsychiatry, 77, 267–281.
Mika, P., Bergner, R. M., & Baum, M. C. (1987). The development of a scale for the assessment of parentifica-
tion. Family Therapy, 14, 229–235.
Schmidt, F. L., & Hunter, J. E. (2003). History, development, evolution, and impact of validity generalization
and meta-analysis methods. In K. R. Murphy (Ed.), Validity generalization: A critical review (pp. 31–65).
Mahwah, NJ: Lawrence Erlbaum.
Ware, J. E., Kosinski, M., & Keller, S. D. (1994). Physical and mental health summary scale: A user’s manual.
Boston, MA: The Health Institute, New England Medical Center.
Ware, J. E., & Sherbourne, C. D. (1992). The MOS 36 Item Short Form Health Survey (SF 36): Conceptual
framework and item selection. Medical Care, 30, 473–483.
Wells, M., Glickauf-Hughes, C., & Jones, R. (1999). Codependency: A grass roots construct’s relationship to
shame-proneness, low self-esteem, and childhood parentification. The American Journal of Family Therapy,
27, 63–71.
Wells, M., & Jones, R. (1998). Relationship among childhood parentification, splitting, and dissociation: Pre-
liminary findings. American Journal of Family Therapy, 26, 331–339.
Wells, M., & Jones, R. (2000). Childhood parentification and shame-proneness: A preliminary study. American
Journal of Family Therapy, 28, 19–27.
Worthington, R. L., & Whitaker, T. A. (2006). Scale development research: A content analysis and recommenda-
tions for best practices. The Counseling Psychologist, 34, 806–838.

14 JOURNAL OF MARITAL AND FAMILY THERAPY