This is an open access article distributed under the terms of the Creative Commons
Attribution License (The central theoretical framework in the field of parental authority deals with describing parenting styles as authoritative, permissive, and authoritarian (Baumrind, 1991; Maccoby & Martin, 1983), which are distinguished by the extent to which the parent sets limits and provides guidance, explains and justifies demands and expectations, employs control and power, and also provides emotional support (Yaffe, 2017a). The fundamental parental dimensions composing these styles of parenting are control/demandingness and acceptance/responsiveness (Darling & Steinberg, 1993; Maccoby & Martin, 1983; Steinberg, 2001). Behavioral control is parentally manifested by setting limits and monitoring the child’s behavior alongside granting autonomy (Steinberg, Lamborn, Dornbusch, & Darling, 1992), while negative parental control (also known as psychological control) is underlined by exaggerated patterns of discipline toward the child, such as overregulation, over-overprotection, and usage of harsh discipline methods (Barber, 1996; Steinberg, Elmen, & Mounts, 1989). Parental acceptance is characterized by aspects of warmth, responsiveness, and additional parental behaviors, such as active listening and giving praise (Maccoby, 1992). The authoritative parent integrates consistent discipline and limit-setting with the provision of warmth and emotional support. Contrary to the permissive parent, who encourages emotional closeness and allows freedom of action to a child, the authoritarian parent is characterized by a high level of psychological control, strict discipline, and a low degree of support and emotional availability (Baumrind, 1968, 1971, 1978; Yaffe, 2017b).
While western individualistic societies have tended to embrace a more democratic parental authority in the last decades (i.e., Authoritative parenting styles; Campbell & Gilmore, 2007), many conservative and collectivistic groups still retain more Authoritarian parenting styles and practices in the family (Dwairy, 2010). Thus, Arab families in the Middle East have been described as emphasizing traditional conservative values, such as obedience to authority and autocratic parenting decision making (Al-Simadi & Atoum, 2000).
Over the years, many measurement instruments have been developed for the self-reporting of children and parents, designed to evaluate the parent’s use of specific practices, to characterize the level of parental control and acceptance, and to classify the parent into one of four parenting styles: authoritative, authoritarian, permissive, and uninvolved (e.g., PPS: Bloom, 1985; PAQ: Buri, 1991; PCRQ: Furman & Giberson, 1995; PAC: Reitzle, Winkler Metzke, & Steinhausen, 2001; PSDQ: Robinson, Mandleco, Olsen, & Hart, 2001; PARQ: Rohner, 2005; CRPBI: Schaefer, 1965; APQ: Shelton, Frick, & Wootton, 1996).
Two of the most popular and widely used instruments to measure Baumrind’s (1971) three basic styles of parenting are Buri’s Parental Authority Questionnaire (PAQ: Buri, 1991) and the Parental Styles and Dimension Questionnaire (PSDQ: Robinson et al., 2001). While both instruments measure the overall parental constructs (e.g., authoritative parenting), the advantage of the latter is its designation to measuring ingredients of the parental dimension described above that theoretically define each parenting styles (i.e., autonomy granting, physical coercion, etc.). In recent years, these two instruments have been increasingly translated and adapted by researchers from all over the world, and have shown a good compatibility for usage in different cultures such as Chinese, Lithuanian, Turkish, Russian, Persian and more (Kern & Jonyniene, 2012; Kiliçgün & Oktay, 2012; Morowatisharifabad et al., 2016; Olivari, Tagliabue, & Confalonieri, 2013). Whereas the Hebrew version of the PAQ has been vastly used in studies in Israel among both the Arab and the Jewish population, to the author’s knowledge only one study has utilized the PSDQ (in its long version) in Hebrew-speaking populations (see: Olivari et al., 2013; Slone, Shechner, & Farah, 2012). The scales of that Hebrew version of the PSDQ in this study have demonstrated good internal consistency reliability for mothers’ and fathers’ reports, averaged together.
Due to the increasing popularity of the PSDQ’s short form (Morowatisharifabad et al., 2016; Olivari et al., 2013) and the lack of a validated Hebrew version, this study was conducted to determine the convergent validity and reliability of the Hebrew-translated 32 items of the instrument among Israeli-Arab families. This short-form exhibited a good reliability data throughout a series of international studies (Olivari et al., 2013), and recently had been most validated cross-culturally (e.g., Kern & Jonyniene, 2012; Morowatisharifabad et al., 2016; Önder & Gülay, 2009). Its adequate psychometric properties in various languages generally support the developers’ claim regarding the questionnaire suitability for multicultural settings (Robinson et al., 2001), and warrants the usage of the short-form (32 items) for further validation in the Hebrew language. Testing the validity of the Hebrew version of the PSDQ in an Israeli-Arab population seemed to be the best choice, since its use by a group in which Hebrew constitutes its second language guarantees the questionnaire’s linguistic-compatibility for most literate Hebrew speaker. The cultural considerations and limitations, however, will be discussed as part of drawing conclusions regarding the measure’s validity. While many other cross-cultural adaptations of the PSDQ have predominantly studied school and middle school-aged children, the current work focuses mainly on adolescents, attempting to test the Hebrew version of its 32-items form in a larger range of ages.
Method
Participants
The sample contained 187 Israeli-Arab family dyads, adolescents (116 boys; 64 girls; 7 participants did not report their sex) and one of their parents (106 fathers; 81 mothers), from several Arab localities in northern Israel. This is a convenience sample, whereby participants were selected by suitability to the research requirements (see the procedure section below), subject to their consent and their literacy skills in Hebrew. The educational distribution of the adolescents’ sample ranged from fourth grade to twelfth grade, with 25% in elementary school, 21% in middle school, and the rest (47%) in high school (7% of the sample did not report their grade). All participants (i.e., parents and adolescents) were skilled in Hebrew-language literacy, qualified to fill in the Hebrew version questionnaires. The children’s age distributed as follows: 28.9% ranged from 10-12, 33.1% ranged from 13-15, and the rest (38%) ranged from 16-18. Finally, parental age was broadly distributed, with 39.5% ranging from 23-40, 57.8% ranging from 41-60, and 2.7% ranging from 61-69.
Table 1 summarizes the demographic data for the parents-children sample.
Table 1
Descriptive Statistics for the Sample’s Demographics (N = 187)
| Variable | Min | Max | M | SD |
|---|---|---|---|---|
| Family number of rooms | 2 | 7 | 3.90 | 0.98 |
| Family size (number of children) | 1 | 8 | 3.90 | 1.28 |
| Adolescent’s age | 10 | 18 | 14.43 | 2.53 |
| Parental age | 23 | 69 | 43.66 | 8.48 |
| Parental education (in years) | 6 | 20 | 13.80 | 2.90 |
Instruments
Parental Authority Questionnaire (PAQ)
The parental authority questionnaire (PAQ: Buri, 1991) was used in the current study as an index for the purpose of testing the convergent validity of the PSDQ’s Hebrew version. The PAQ contains 30 items and is used to classify parents into to one of Baumrind’s three parenting styles conceptualization (Baumrind, 1971), based on the child’s self-report: Authoritative (10 items, e.g., "As I was growing up, once family policy had been established, my parents discussed the reasoning behind the policy with the children in the family"), Authoritarian (10 items, e.g., "As I was growing up my parents did not allow me to question any decision they had made"), and Permissive (10 items, e.g., "As I was growing up my parents seldom gave me expectations and guidelines for my behavior"). The response scales for an item range from 1 (strongly disagree) to 5 (strongly agree). The index for each parenting style is the sum of the relevant items of each scale. Thus, the total score for each parenting scale ranges from 10 to 50, and a higher score reflects a higher specification of the style. It is a valid questionnaire with a relatively high internal reliability and test-retest reliability (0.74 to 0.78) (see: Buri, 1991; Smetana, 1995), widely used in Israel (e.g., Enten & Golan, 2009; Mayseless, Scharf, & Sholt, 2003; Yaffe, 2017b, 2018) and around the world to measure Baumrind’s (1971) three basic styles of parenting. Previous research supported the validity of the PAQ’s Hebrew version and had shown adequate rates of reliability for it (Yaffe, 2018). In the current study, we recorded an adequate internal consistency for the three scales of the Hebrew measure (see Alpha coefficients in Table 2), which is consistent with the reliability data found for the tool in past research.
Table 2
Means, Standard Deviations and Alpha Coefficients of the Parental Questionnaires
| Scale | M | SD | Item Numbers | α |
|---|---|---|---|---|
| PAQ | ||||
| Permissive | 31.68 | 6.32 | 1, 6, 10, 13, 14, 17, 19, 21, 24, 28 | .72 |
| Authoritarian | 33.50 | 6.17 | 2, 3, 7, 9, 12, 16, 18, 25, 26, 29 | .71 |
| Authoritative | 36.24 | 6.12 | 4, 5, 8, 11, 15, 20, 22, 23, 27, 30 | .64 |
| PSDQ | ||||
| Permissive (Indulgent) | 2.68 | 0.66 | 17, 20, 8, 15, 24 | .52 |
| Authoritarian | 2.39 | 0.58 | (12 Items) | .79 |
| Physical Coercion | 2.12 | 0.64 | 2, 6, 19, 32 | .53 |
| Non-Reasoning/Punitive | 2.40 | 0.79 | 4, 10, 26, 28 | .62 |
| Verbal hostility | 2.65 | 0.81 | 13, 16, 23, 30 | .67 |
| Authoritative | 3.92 | 0.61 | (15 Items) | .89 |
| Warmth & Support | 4.10 | 0.70 | 7, 1, 12, 14, 27 | .78 |
| Autonomy granting | 3.62 | 0.70 | 3, 9, 18, 21, 22 | .70 |
| Regulation | 3.93 | 0.66 | 25, 31, 11, 5, 29 | .76 |
Note. N = 187.
Parenting Styles and Dimensions Questionnaire (PSDQ)
The parenting styles and dimensions questionnaire (PSDQ: Robinson et al., 2001), originally measures the usage of various parental practices in adolescents’ parents, through which the parent can be classified into one of the three parenting styles according to Baumrind's typology (1991). The PSDQ with 32 items used here is a modified version of the original 62-item PSDQ (also known as the Parenting Practices Questionnaire – PPQ: Robinson, Mandleco, Frost Olsen, & Hart, 1995), originally adjusted by the authors for use with parents of children from 4 to 12 years old in various cultures. The 32-items form of the PSDQ was developed using Structural Equation Modeling on 1900 mothers and fathers of preschool and school-aged children (Robinson et al., 2001), and was widely used around the world in studies on parents of school and middle school-aged children (Olivari et al., 2013). It contains three overall scales of parenting styles: Authoritative (15 items - e.g., “I give our child reasons why rules should be obeyed”), Authoritarian (12 items - e.g., “I scold or criticize when our child’s behavior doesn’t meet our expectations”), and Permissive (5 items - e.g., “I spoil our child"). The Authoritative scale reflects the three dimensions of warmth/support (e.g., “I encourage the child to talk about the child’s problems”), regulation (e.g., “I emphasize the reasons for rules”), and autonomy granting (e.g., “I allow the child to give input into family rules”). The Authoritarian scale comprises three dimensions of physical coercion (e.g., “I slap my child when the child misbehaves”), verbal hostility (e.g., “I explode in anger toward our child”), and nonreasoning/punitive strategies (e.g., “I use threats as punishment with little or no justification”). The Permissive scale has 5 items assessing the sub-factor of indulgence. The questionnaire is designed for the parent’s self-report and/or his/her spouse, and in its measured on a 5-level Likert scale (1 = never 5 = always). The measure yields a separate, continuous score for each dimension of parenting. The index for each parenting dimension and style constitute the mean score of the relevant items on each scale, with larger numbers indicating a broader use of parenting practices associated with a particular style. The English version of the PSDQ has adequate internal consistency reliability. Robinson et al. (2001) reported the Cronbach’s for mothers’ and fathers’ report to be .86, .82, and .64 for Authoritative, Authoritarian, and Permissive Parenting scales, respectively.
The PSDQ is considered as one of the few instruments available with psychometrically defensible scales relating to parenting practices and styles assessment (Locke & Prinz, 2002). Although this measure is widely used in the literature and has been adapted for effective use in multiple cultural settings, its usage in Hebrew-speaking populations was surprisingly meager. The only known study using Hebrew translated form of this measure in Hebrew-speaking families (mentioned above; Slone et al., 2012) yielded a good Cronbach’s alpha coefficients for the authoritative scale (.85 and .84), for the authoritarian scale (.84 and .81), and for the permissive scale (.79 and .81) for mothers and fathers respectively.
Procedure
The translation of the original instrument was carried out following the procedure of back–forward translation according to recommendations prepared by Van de Vijver and Hambleton (1996). The author translated and adapted the instrument’s original form from English into Hebrew (the self-reporting version only), and the translated items were reviewed and translated back to English by the author’s trilingual independent colleague (French, English, and Hebrew speaker), who did not take part in the current study. This was followed by a bilingual (Hebrew and English speaker) professional translator and language editor in English who translated the instrument back into the Hebrew language. All three translations were reviewed and analyzed by an educational psychologist whose native language was Hebrew. Prior to conducting the study, the completed Hebrew version of the PSDQ was administrated to 5 Israeli-Arab Hebrew speakers (all of them parents), who completed the questionnaire without reporting any language difficulties.
As part of the data collection, two of the study’s research assistants visited several Arab localities in northern Israel one after the other for about three months. In each locality, they met with families (mother/father and adolescent) in a neutral place or at the family's home and instructed them to fill-in the questionnaires. The participants were given strict guidance regarding the procedure for completing the instruments, after they were informed of the research objectives and their ethical rights were clarified. The research instruments were administered anonymously. When filling out the questionnaires, parents and children were asked (both verbally and in writing) to refer to their parent/child who was filling out the other questionnaire (i.e., the family member who took part in the study). An ad-hoc institutional research committee discussed and approved the research process prior to its implementation.
Results
First, we display and discuss the measures’ scores (means and standard deviations). Subsequently, we examine the indicators of reliability and validity obtained for the PSDQ scales in its Hebrew version. Table 2 presents the means and standard deviations recorded in the current sample for both PSDQ and the PAQ (these scores did not significantly differ between genders, therefore the data is displayed for the sample as a whole). The mean scores for the PSDQ subscales range between 2.12 (physical coercion) and 4.10 (warmth & support), while the mean scores for the PSDQ overall parenting scales range between 2.68 (permissive) and 3.92 (authoritative).
Consistently, both measures recorded a higher mean score for the authoritative parenting scales, compared to the non-authoritative parenting scales. These differences are significant either for the PAQ using the child’s report, F(2, 185) = 32.11, p < .001, and for the PSDQ using the parental report, F(2, 185) = 275.41, p < .001. That is to say, both the parents and their children perceive the sample’s parents as more authoritative in their style than permissive and authoritarian. From a methodologic perspective, this indicates agreement between the two measures.
Reliability and Convergent Validity
In order to assess the reliability of the PSDQ in its current version, we calculated the Cronbach’s alpha coefficients for the measure’s overall scales and subscales. The results of these tests are shown in Table 2, indicating adequate internal consistency indices (with two exceptions), which range from .52 to .89 for the overall scales and from .53 to .78 for the subscales. The two scales whose Alphas are below .60 (that is, permissive and physical coercion) might be partially a result of the small number of items they contain, although their properties should be accordingly accounted.
To determine the convergent validity of the PSDQ in its current Hebrew version, we calculated the Bivariate correlations between the measure’s scales and the PAQ’s three parenting styles scales (Table 3), using different informants for each measure (i.e., parent reports on himself / herself via PSDQ, and child reports on parent via PAQ). Convergent validity of an instrument is manifested by a statistical association between its scores and the scores of a separate assessment of the same theoretical construct (Campbell & Fiske, 1959). Cohen (1988) provided general guidelines for determining the magnitude of correlations, where those between 0.10 and 0.29 are considered small, between 0.30 and 0.49 are moderate, between 0.50 and 0.69 are large, and between 0.70 and 0.90 are very large. In terms of these threshold levels, we found moderate to large significant correlations between the corresponding overall scales of parenting styles in the two questionnaires, indicating a general convergence between the measurements. All of these correlations are significant at 0.1%, and remain steady across gender and after controlling for SES and parents’ years of education. Underlying these trends are the PSDQ subscales’ convergence against the corresponding PAQ overall scales, with correlations ranging from a moderate to an approximate large magnitude. Taken together, these results establish the construct validity of the PSDQ in its current Hebrew version, due to its convergence with a separate valid index of parenting styles.
Table 3
Correlation Coefficients Between the PSDQ Scales and the PAQ scales (N = 187)
| PSDQ (parental report) | PAQ (child’s report)
|
||
|---|---|---|---|
| Authoritative | Authoritarian | Permissive | |
| Authoritative | .53** | -.06 | .01 |
| Warmth & Support | .49** | .00 | .04 |
| Autonomy granting | .43** | -.04 | .12 |
| Regulation | .46** | -.12 | -.10 |
| Authoritarian | -.01 | .47** | .36** |
| Non-Reasoning/Punitive | -.08 | .49** | .44** |
| Physical Coercion | -.14 | .36** | .31** |
| Verbal hostility | .15* | .30** | .12 |
| Permissive (Indulgent) | .04 | .44** | .48** |
*p < .05. **p < .001.
We also discovered unexpected positive correlations between the permissive and the authoritarian scales in both directions (that is, the permissive scale of both measures was correlated with the authoritarian scale of the other measure). Similar association trends with a larger magnitude were also recorded between these scales internally for each questionnaire (PAQ: r = .55, p < .001; PSDQ: r = .59, p < .001), which signifies a correspondence between the two measures even regarding anomaly trends. Nevertheless, with a small-insignificant range of correlations, the results showed good discrimination between the PSDQ’s authoritative scale (and its subscales) and the non-authoritative scales of the PAQ measure.
Discussion
This study strived to identify the level of reliability and to provide information on the convergent validity of the Hebrew version of the PSDQ, when administered to Israeli-Arab parents of adolescents whose age ranges from 10 to 18. Also, it provides the basic psychometric properties that are comparable to the original version of the instrument. Despite its popularity among researchers from all over the world, the Hebrew version of the PSDQ has seldom been used in Hebrew-speaking populations and had never been psychometrically tested before.
The scores obtained in the current sample (means and standard deviations) for the PSDQ scales in its Hebrew version generally correspond to the norms recorded by the instrument’s developers (Robinson et al., 2001). Also, the mean scores recorded for the overall parenting scales are essentially similar to those reported for the original measure (subject to slight gaps, which can be partially attributed to cultural differences), with the same hierarchical score between them (i.e., authoritative, authoritarian, and permissive in descending order). This order was also observed when using the PAQ instrument with the adolescents’ report, indicating the agreement between the measures regarding the sample’s parents as being more authoritative and, thereby, reinforcing the validity of the PSDQ in its current version.
As for the reliability results, apart from two exceptions, all alpha coefficients obtained in the current study were above .60, with the majority of them ≥ .70. This generally indicates an acceptable internal consistency data, which meets the minimal reliability requirements for using the current PSDQ measure in its Hebrew version. The study reviewed in the introduction section using the Hebrew version of the long-form PSDQ (Slone et al., 2012) reported averagely higher coefficients for the Authoritarian and the Permissive scales (.82 and .80, respectively), which can be attributed partially to their greater number of items. Moreover, as compared to the results obtained in the current work for the Hebrew version, the English version of PSDQ reported quite similar internal consistency data, with Cronbach’s coefficients of .86, .82, and .64 for authoritative, authoritarian, and permissive parenting scales, respectively (Robinson et al., 2001). Our datum regarding the permissive scale is slightly lower than reported for the original measure, yet this scale had shown consistent lower reliability data across several studies. In their comprehensive review of the psychometric properties of the PSDQ, Olivari et al. (2013) concluded that the permissive parenting style appears to be less reliable than the other two styles, partially due to the relatively lower number of items it contains. They found Cronbach’s alpha levels to be generally adequate for the authoritarian (.62-.95) and the authoritative (.71-.97) scales, with consistently lower reliability for the permissive scale (.38-.95).
Another subscale to show an alpha level below .60 was the physical coercion scale. This four-item subscale as part of the authoritarian overall parenting style has previously exhibited slightly higher, mostly acceptable, reliability levels in various cultural settings (Kern & Jonyniene, 2012; Olivari et al., 2013; Önder & Gülay, 2009). Thus, this scale should be further monitored and inspected in additional samples when using the Hebrew version of the PSDQ.
The main aim of the current study was to evaluate the convergent validity of the current version of PSDQ. Convergent validity denotes the extent to which an instrument’s output is associated with other assessments’ output intended to measure the same construct (Ayán-Pérez, Martínez-Lemos, & Cancela-Carral, 2017). Campbell and Fiske (1959) stated that evidence of a convergent validity of an instrument is manifested by agreement between its scores and the scores of distinguished-independent assessments of the same theoretical construct. Adhering to this instruction, we used different reports on separate measures of parenting styles as an index for the Hebrew version of the PSDQ.
The pattern of correlations recorded in the current sample indicated convergence between the PSDQ and the PAQ instruments, with all coefficients exceeding the magnitude of |0.35| suggested by Hammill, Brown, and Bryant (1989) as the threshold correlation to constitute evidence of convergent validity. The moderate to high correlations obtained between the theoretically related and overlapping constructs measured by the two instruments indicate the existence of a good convergent validity for the PSDQ in its Hebrew version. These results remained steady across gender and even after controlling for SES and parents’ education. Taken together with the reliability results, these findings are generally supportive of Robinson’s et al. (2001) claim that the PSDQ can be used in multicultural settings, and join previous evidence supporting the measures’ multilanguage construct validity in different cultures (see: Morowatisharifabad et al., 2016; Olivari et al., 2013).
Having said this, further research should be conducted in a larger sample of Hebrew-speaking populations (including Jewish parents) to supplement the lacking psychometric properties of the PSDQ in its current version (e.g., factorial analysis). Moreover, the associations documented here between the permissive and the authoritarian scales (as they were significantly positively correlated), raises some concern regarding the instruments’ lack of discrimination between these contrasted parental constructs in some cultural contexts. One possible interpretation suggested by Olivari et al. (2013) in this regard is that the meaning of the permissive-style items is interpreted differently in different cultural contexts. This assumption is consistent with the repetitive association found here between the permissive and authoritarian scales, both within and between the two measures (i.e., between the PAQ scales, between the PSDQ scales, and between the PAQ and the PSDQ scales). Our findings indeed showed a good discrimination between the PSDQ authoritative scale (and its subscales) and both the PSDQ and the PAQ non-authoritative scales. Yet, a separate inquiry in a culturally distinguished Hebrew-speaking population (such as Jewish families) is required to determine the instrument’s ability to fully reflect the theoretical relations between the parental constructs.
Conclusions and Limitations
The study tested the convergent validity and reliability of the Hebrew version of the PSDQ in Israeli-Arab families. Its strength derives from the method used to validate the instrument against a separate assessment of parenting styles, while using different informants (i.e., adolescents) and distinct index (i.e., PAQ). The PSDQ demonstrated a good convergence with the relevant PAQ scales, and met the minimal psychometric requirements for using its Hebrew version. Yet, these results are limited in several aspects, apart from the limited sample size the study uses (as aforesaid). First, the current sample consisted of educated parents, which means that the Hebrew version of this instrument may not suit Israeli-Arab parents with lower educational levels, whose Hebrew literacy skills are basic. Moreover, while the gender division of the parental sample was reasonable, due to sampling constraints the adolescents’ sample suffered from a significant imbalanced partition, with boys approximately as twice frequent than girls. This might compromise the instrument’s validity with reference to the child’s gender. Hence, the psychometric properties of the current Hebrew form of the PSDQ required further examination in larger and gender-balanced samples. Finally, in order to address several cultural and psychometric issues discussed above in relation to the instrument’s validity, a similar method should also be employed in Israeli-Jewish families using the Hebrew version of the PSDQ.