Jouiima] of Family Psychology 2014, 14, Vol. 28, No. 3, 401-406
© 2014 American Psychological Association 0893-3200/14/S12.00 DOI: 10.1037/a0036370
BRIEF REPORT
Eavesdropping on the Family: A Pilot Investigation of Corporal Punishment in the Home George W. Holden, Paul A. Williamson, and Grant W. O. Holland Southern Methodist University This study tested the feasibility of using audio recorders to collect novel information about family interactions. Research into corporal punishment (CP) has relied, almost exclusively, on self-report data; audio recordings have the promise of revealing new insights into the use and immediate consequences of CP, So we eould hear how parents respond to child conflicts, 33 mothers wore digital audio recorders for up to 6 evenings. We identified a total of 41 CP incidents, in 15 families and involving 22 parent-child dyads. These incidents were evaluated on 6 guidelines culled from the writings of CP advocates. The results indicated, contrary to advice, CP was not being used in line with 3 of the 6 recommendations and for 2 others, the results were equivocal. The last recommendation could not be assessed with audio. Latency analyses revealed children, after being hit, were misbehaving again within 10 minutes after 73% of the incidents. Mothers' self reports about whether they used CP were found to correspond to the audio data in 81% of the cases. Among the mothers who were hitting, CP occurred at a much higher rate than the literature indicates. These results should be viewed as preliminary because of the small sample of families and the even smaller number of families who used CP. Nevertheless, this pilot study demonstrates that audio recording naturally occurring momentary processes in the family is a viable method for collecting new data to address important questions about family interactions. Keywords: audio recordings, corporal punishment, discipline, ecological momentary assessment, home observations
The overreliance on behavioral self-reports is a well-recognized limitation of contemporary research (e,g,, Baumeister, Vohs, & Funder, 2007), Self-reports are limited by recall abilities, lack of awareness, social desirability, and other sources of distortion (Morsbach & Prinz, 2006), Self-reports also limit the types of questions that can be addressed, such as with regard to sequential analyses or latencies. Observational data, though it has problems of its own (e,g,, Gardner, 2000; Hops, Davis, Longoria, 1995), is able to capture the sequence of actual ongoing behavior. In this study, we report on the feasibility of using audio recordings to provide new insights into one way in which parents respond to child misbehavior. Although audio recording family interactions is not a new methodology (e,g„ Johnson & Bolstad, 1975), it is rarely used. The advantage of audio recorders over video recorders is that because a camera operator is not needed, audio devices use
fewer resources and lessen the likelihood of reactivity (Gardner, 2000), Despite the absence of visual data, audio provides rich interaction data and can ufilize the ecological momentary assessment approach, characterized by repeated sampling of subjects in real fime in their natural environments (Shiffman, Stone, & Hufford, 2008). We used audio recordings to investigate how parents use corporal punishment (CP). This is a prime topic to study because the extensive literature on the topic relies almost exclusively on selfreports. This reliance on one method of data collection limits our understanding of CP and its impact. For example, parents may distort their reports (consciously or otherwise), they may be unaware of their own behavior, and they are likely unable to report, with fidelity, parent-child interaction sequences. It is also an important topic to investigate given the intemational debate over the use of corporal punishment (e,g,, Benjet & Kazdin, 2003; Durrant & Ensom, 2012), Despite its widespread use both in the United States and internationally (Gershoff et al, 2010; Lansford, Wager, Bates, Dodge, &. Pettit, 2012), opponents of CP argue that spanking or slapping is ineffective, and cite the large number of studies that link it to a variety of unintended child behavior problems and other negative sequelae (e,g„ Durrant & Ensom, 2012; Gershoff, 2013; MaguireJack, Gromoske, & Berger, 2012), On the other hand, proponents of CP are critical of the corpus of research, arguing that the data are correlational and have various methodological limitations, such as not controlling for child effects (e,g,, Larzelere, 2000), Advo-
This article was published Online First April 14, 2014, George W. Holden, Paul A. Williamson, and Grant W. O, Holland, Department of Psychology, Soudiem Methodist University. We thank the numerous research assistants who aided in the collection and coding of data. This research was supported by a grant from the Timberlawn Psychiatric Research Foundation. Correspondence concerning this article should be addressed to George Holden, Department of Psychology, Southern Methodist University, P.O. Box 750442, Dallas, TX 75275-0442. E-mail: gholden@smu,edu 401
HOLDEN, WILLIAMSON, AND HOLLAND
402
cates also point out that some studies "lump together" all types of corporal punishment, thereby conflating the effects of abusive practices with "normative" spanking or "customary" spanking. They hold that the effects of spanking depend on how it is used, and if used appropriately, it can be an effective form of power assertion (Baumrind, 1997, 2012). Audio recordings of family conflicts can help to fill in a major gap in the CP literature—that of the immediate and concomitant consequences of CP (Benjet & Kazdin, 2003). Despite the hundreds of studies investigating the associations between CP and child behavior, there continues to be a paucity of information about what transpires before, during, and immediately after disciplinary incidents involving CP. As a result, it is impossible to evaluate the validity of proponents' assertions that a) there is a "customary" form of CP, b) that customary CP follows suggested guidehnes, and c) that this form of CP is effective for disciplining children. Assessing the validity of these assertions requires naturalistic observations. To date, observations of parents engaging in CP are extremely rare. Brown (1979) observed a total of 20 "hitting or physical force" responses in a shopping mall. In a study of verbal aggression, Davis (1996) reported that he observed parents hitting their children in public, backing up threats of hitting about one third of the time. The Home Observation for the Measurement of the Environment (e.g., Bradley, Corwyn, McAdoo, & Garcia Coll, 2001) measures CP simply as present or absent while the interviewer is present. Recently, Stansbury, Haley, Lee and BrophyHerb (2012) observed "negative touches" (slapping, pinching, hitting, or threatening to hit) in response to noncompliance in public settings. In 23% of the incidents, children received negative touches. The reports of these observational studies do not provide enough information to determine whether the parents followed prescribed recommendations or whether the CP was effective in the immediate context. Nor do they shed light on the proximal determinants of CP, such as the eliciting misdeed or parental affect (Ateah & Durrant, 2005). In the present study, we use home observations to examine the use of CP and to assess its immediate impact on children's behavior. We evaluate how it is used in light of "best practices" guidelines provided by proponents. Seven guidelines were identified; a) it should be used infrequently; b) it should be used selectively, for serious misbehaviors, such as aggression; c) it should be used as a last resort; d) it should be administered calmly, not in anger; e) it should consist of no more than two hits; f) the child should be struck on the buttocks; and f) it should be painful. These guidelines, gleaned by Vittmp, Holden, and Buck (2006), came from five sources (e.g., Dobson, 1992; Rosemond, 1994; Trumbull & Ravenel, 1998). In sum, we sought to explore whether home audio recordings could reveal new information about CP use in the home and whether it adhered to the parameters recommended by proponents of the disciplinary technique. We also investigated the underlying assumption of proponents that CP is effective at curtailing subsequent misbehavior. Einally, we addressed a methodological question: Do self-reports accurately reflect the observed frequency of CP?
Method The study complied with the APA's ethical principles and was approved by the university's IRB committee.
Participants Participants were recmited using flyers distributed to Englishspeaking mothers of 2- to 5-year-old children at daycare and Head Start centers in a large southwestem city. A total of 56 potential participants contacted the investigators. Only the mothers who reported, in a phone screen about disciplinary practices, that they yelled in anger at least twice a week (the original study focus) were invited to participate. Eourteen mothers did not meet this criterion. In addition, seven mothers dropped out. Attrition analyses on the screening data showed the drop-outs did not differ (race/ethnicity, number of children, or frequency of spanking) from the participants. Thirty-five mothers of 2- to 5-year-old children participated. Audio recordings from two families were insufficiently intelligible to use. In the final sample of 33 families, the target children averaged 46.0 months of age (SD = 11.8, range 24 to 68); 13 were female. The median number of siblings was one (range 0 to 3). The mothers averaged 33.8 years old (SD = 5.8, range 23 to 46), most were married (82.1 %), and self-reported as white (60.7%), African American (25.0%), or Hispanic (10.7%). The mothers were employed full-time outside the home (60.7%), worked part-time (17.9%), were homemakers (14.3%), or were students (3.6%). One third of the mothers (32.1%) had attended some college or vocational school, another third (32.1%) were college graduates, and 28.6% had advanced degrees or training. Eathers' average age was 35.7 years (SD = 4.5) and most (91%) were employed full time. Their education ranged from high school (14%) to graduate or professional school (24%). Annual family income included less than $40,000 (28.6%), between $40,000 and $79,999 (32.2%), $80,000 or more (39.2%).
Measures Audio recordings. Mothers, visited at their homes, were given an Olympus DS-50 digital voice recorder to wear, in a sport pouch, on the upper arm. They were instmcted to turn on the recorder upon returning home from work (or at 5:00 p.m.) and turn it off once their child fell asleep. The first 10 participants were asked to make recordings on four consecutive days; the remaining participants were asked to record six consecutive days. This procedure resulted in an average of 12.95 hrs of recordings per family (SD = 6.16, range 2.1 to 25.5). Most (94%) of the mothers complied with the number of recordings requested, though only 62% of the recordings continued all the way through bedtime. When the recordings were completed, a researcher retumed to the home to collect the audio recorder and questionnaires. Mothers received $120 for their efforts. The first step in coding required identifying the CP incidents. In 51% of the incidents, the sound of the child being slapped or spanked was clearly discernible and supported by contextual cues, such as wamings of or justifications for the hit. In 44% of the incidents, the sound of CP was ambiguous, but the contextual cues (mother's warning of impending CP, child's cries) provided supporting evidence. In two cases (5%), there was no audible sound of CP but clear explicit contextual information, such as the child pleading, "Stop hitting me." Two raters independently coded 30% of the recordings. The interrater reliability, computed by agreements over agreements and disagreements, was .87, with a Co-
EAVESDROPPING ON THE FAMILY hen's kappa (K) of .77. Disagreements for all codes were discussed and settled by consensus. The second step involved a detailed analysis by two coders of each incident relating to the seven guidelines. However, given that we were relying on audio data, the location on the body as to where the child was struck could not be determined. Therefore, that guideline was not evaluated. We derived the dependent variables for the other six guidelines in the following ways: CP should be used infrequently. To measure the frequency of CP use, we calculated the rate of CP per hour of recording. Because only mothers wore the recorders, this analysis was limited to them. The interrater reliability was .87, K = .77. CP should be used selectively, for serious behaviors. To assess the types of behaviors eliciting CP, we coded the child transgressions into five categories: a) social convention (e.g., personal untidiness or uncleanliness, poor manners); b) aggression (verbal or physical acts that could hurt someone); c) rights violation (e.g., taking someone's property; intruding on other's personal space); d) destruction (intentional acts that could damage objects); e) prudential rule violation (engaging in a safety risk such as touching a stove). The first four categories came from Dunn and Munn (1987). All but two of the CP incidents could be coded into one of these categories. The reliability was .73, K = .58. CP should be used as a last resort. This guideline was operationalized in two ways. First, we measured the latency in seconds from the onset of the child's infraction until the CP. The intraclass correlation between the two raters was .71. Second, we summed the number of responses to the transgression prior to the first hit. The agreement was .86; ICC = .44. CP should not be used in anger. We assessed, on a dichotomous code, the presence of parental negative affect when the CP was delivered. Negative affect occurred if the volume of the utterance increased, the tone of voice indicated aggravation or anger, and/or the presence of negative verbalizations (e.g., threats). The interrater reliability was M = .73, K = .46. The CP should consist of no more than two hits. Based on the sound or verbal comments, we counted the number of slaps/ spanks. The reliability was very high, ICC = .99. The hit should be painful. We operationalized the degree of pain experienced by the child on a 4-point distress scale (0 = no audible response; 1 = minimal but audible, such as whimpering; 2 = moderate, may include some crying or verbal protest; 3 = strong, such as loud crying). The interrater reliability was .71, K = .53. "Appropriate" CP is effective. Effectiveness of CP was operationalized as whether a new conflict occurred within 10 min of the CP. The agreement (within 5 s) was .90, K = .78.
Questionnaires To assess the correspondence between self-reported CP and observed frequency of CP, we administered two questionnaires. Each mother completed the PRCM twice: once at the time of screening (Time 1) and once at home during the period when recordings were being made (Time 2). The PSDQ was only filled out at Time 2. Parental Responses to Child Misbehavior (PRCM). The PRCM (Holden et al., 1995) asks parents how frequently, in an average week, they use particular disciplinary responses, rated on
403
7-point scales (0 = Never, 1 = less than once; 2 = 1-2 times; 3 = 3-4 times; 4 = 5-6 times; 5 = 7-8 times; 6 = 9 or more times). The original PRCM has good test-retest reliability, internal consistency, and validity (e.g.. Holden et al., 1995; Vittrup et al., 2006). The original PRCM contained only one item referring to CP ("spank/slap"). For the present study, we revised that item to "spank with hand" and added three more CP items: "spank with an object," "slap on the child's hand," and "slap on the child's face." However, the last item lowered the internal consistency of the new four-item CP subscale. Consequently we removed it. The final 3-item PRCM-CP subscale correlated strongly across Time 1 and Time 2 (r = .75, p < .001), but was only moderately internally consistent (a^ime i = -52, a.rime 2 = -61). Parenting Styles and Dimensions Questionnaire (PSDQ). The PSDQ (Robinson, Mandleco, Olsen, & Hart, 1995) is a commonly used 62-item measure. For the purposes of this study we report only on the six-item CP subscale (e.g., "I spank when our child is disobedient"), rated on a 5-point frequency scale (1 = Never; 5 = Always), with an a of .70 in this sample.
Results CP Incidents A total of 41 CP incidents were heard in 15 of the 33 families (45% of the sample). The distribution of incidents within these families ranged widely. Six families each contributed one incident, four families had two incidents, one family had three incidents, two families had four incidents, one family had six incidents, and the last family had 10 incidents. Only mothers used CP in nine families, only fathers in three families, both parents in two families, and mother and grandmother in one family. (For convenience, we will refer to all these individuals as "parents.") Eighteen children (11 boys) received CP. Their median age was 48 months (range 7 months to 6 years). Twelve mothers accounted for 32 incidents, five fathers for seven incidents, and one grandmother for two incidents. Because some families contained multiple children and two adults, we calculated the total number of parent-child dyads involved. The incidents involved 22 different parent-child dyads. Sixteen of the parent-child dyads had one incident, one dyad experienced two incidents, two dyads had three, and one dyad each contributed four, five, and 10 incidents. Three children were hit by both their parents, and three parents hit two children. The incidents were more frequent in the first three evenings (six, 11, & 10 incidents, respectively) than in the last three (six, one, & seven, respectively).
Evaluating CP Incidents in Light of Proponents' Guidelines Next we summarize the correspondence between the audio data and the six guidelines. CP should be used infrequently. The median rate was .22/hr (SD = .16) with a low of .06 per hr (based on one incident for 17.95 hrs) to a high of 1.43/hr (three incidents from 2.1 hrs). CP should be used selectively, for serious behaviors. To evaluate selectivity in use, we examined the eliciting misbehavior. In 40 of the 41 incidents we identified the misdeed or source of the
404
HOLDEN, WILLIAMSON, AND HOLLAND
confiict. Noncompliance was the proximate cause in 90% of the incidents. Typically, this noncompliance consisted of the child ignoring or resisting parental instruction or command ("Come here!"). The immediate cause of the remaining four incidents was aggression or breaking a standing rule (resulting in CP without warning). The misdeed that precipitated the noncompliance was coded into the five behavioral domains. Seventy-four percent of the transgressions involved breaking social conventions. The offending misbehaviors consisted of such actions as sucking fingers, eating improperly, getting out of a chair, and going outside without permission. The next most frequent category (10%) was acts of aggression, including the child hitting someone or throwing an object at the parent. Rights-related misdeeds occurred in 7.5% of the incidents. The remaining misdeeds were prudential (5%) or destructive (2.5%). CP should be used as a last resort. On average, parents tried one disciplinary response before the CP (M = 1 . 1 , SD = .97, range 0 to 3). Most commonly, the initial reaction was a prohibition (e.g., "Stop it!"). The elapsed time from the onset of a confiict to administration of the CP averaged less than 30 s (M = 26.5, SD = 67.31, range 0 to 416). CP should not be used in anger. In 49% of the incidents, parents expressed negative affect while delivering the CP. The CP should consist of no more than two hits. Only one hit was audible in 83% of the incidents. In three cases (7.3%), two hits were heard. There was one incident each with 3, 5, 6, and 11 hits. The mean number of hits was 1.59 {SD = 1.83). The hit should be painful. The modal child distress rating was moderate (48.8%), followed by minimal (29.3%), and strong negative reaction (9.8%). No audible child reaction was heard in 12.2% of the incidents. The mean distress rating was 1.56 {SD = .84). Finally, to gauge the effectiveness of CP, we coded whether the cMld misbehaved during the immediate 10 tnin that followed the CP. In 30 of the 41 CP incidents (73%), the children engaged in the same or another misbehavior within that time period. To supplement these descriptive statistics, we conducted multilevel analyses in an effort to account for nonindependence of observations. Those analyses did not differ substantially from the results presented here, so are not reported (available from the first author).
Correspondence Between the Audio and Questionnaire Data We examined the correspondence between the two forms of data in two ways. The correlation between the CP subscale on the PRCM (Tl) and the actual number of CP incidents was not significant, r = .21, p = .23. However, when using the CP subscale from the Time 2 (r[30] = .76, p < .001 with Tl), the correlation increased to r = .54, p < .001. That can be accounted for by the tendency for mothers to report more CP at Time 2{M = 2.87) than at Time 1 (M = 2.33), paired r(29) = 1.44, p = .16. The CP subscale from the PSDQ survey also correlated significantly with incidents, r = .47, p < .001, as well as with the PRCM subscale (r^-j^^ , = .58, r.^^^^ ^ = -68, ps < .001). When examining reports of CP dichotomously (spank vs. no spank) based on whether mothers reported they used CP at least 1-2 times per week
on the PRCM (T2), reports and behavior showed good correspondence, x^(l) = 11.79, p = .001. Seventeen mothers reported they did not use CP (or did so less than once/week) and were not heard to use CP. Nine mothers who reported they did use CP, indeed used it. Four mothers reported they did use CP but were not heard to; two mothers reported they did not use CP but used it.
Discussion The use of audio recorders proved to be a highly effective technique to study the moment-to-moment interactions surrounding conflict in the home. The method afforded new data about how "customary" CP was being used in the home—at least by this small sample of parents. The audio recordings allowed us to assess CP use in the home and examine several new variables related to the CP use that had not been previously reported in the literature. The recordings revealed CP occurred, on average, just 30 s after the onset of a conflict and with little warning. These conflicts typically resulted from the violation of a social convention. Parents, clearly angry in half the cases, typically hit their children once or twice per incident. Within 10 min of the physical discipline, in 73% of the events examined, children misbehaved again. The recordings revealed that CP was frequently not being used in the ways that advocates of the disciplinary technique recommended. Three caveats about generalizability are in order. First, this small self-selected sample of parents was screened for yelling and the mothers are more educated than the general population. It is possible these parents used more frequent, harsher disciplinary practices than would be found in a more representative sample. In addition, the recordings were made during the most stressful time period of the day, when emotional spillover from prior events are likely and children are most at-risk for CP (Holden et al, 1995; Matjasko & Feldman, 2006). Therefore, the observed rates of CP would likely be lower if recordings had been made at other times of the day. Third, CP was heard in only 22 parent-child dyads from 15 families. Evidently, this study should be considered a pilot effort and a larger, more representative sample is needed. Of the parents who used CP, their behavior commonly violated three of the six "use" guidelines examined; infrequent, serious misbehavior, and last resort. Among the 12 mothers who engaged in CP, the median rate was once every 6.3 hrs of interactions. In about three-quarters of the incidents, parents hit their children for extraordinarily mundane social convention offenses, rather than serious infractions, such as rights or prudential violations (Dunn & Munn, 1987). Most poignantly, one child was slapped for turning pages of a storybook. Parents used CP as their second, not last, resort, on average just half a minute after the conflict began. With only a few exceptions, parents were abiding with one of the guidelines—hitting only once or twice. The data were equivocal for the last two guidelines. Although the guidelines call for calmly metering out punishment, parents were rated as angry in about half of the incidents. Consequently, these data suggest that the role of negative emotion, while present in many incidents, should not be exaggerated as a determinant (Ateah & Durrant, 2005). The final guideline (making CP painful) was indirectly assessed with a rating how upset the child sounded. In just over half of the cases, the children were heard to be at least moderately
EAVESDROPPING ON THE FAMILY
upset after the CP. Videos would provide better assessments of children's emotional response. When all six guidelines are considered together in one incident, it is clear that parents were rarely, if ever, using CP "appropriately." The recordings reveal that most parents were responding either impulsively and/or emotionally, rather than instrumentally and intentionally (Benjet & Kazdin, 2003). Impulsivity may have attenuated the correlation between self-reports of CP and observed incidents. Nevertheless, at Time 2, when the mothers were somewhat more forthcoming about their CP behavior, the x^ test was significant for the reports of and overheard CP incidents, matching in 81% of the cases. Most of the disagreements (12%) involved mothers who reported they used CP but were not heard to. This discrepancy is not surprising given the recordings averaged about 13 hrs over four to six days and mothers were reporting on behavior across an average week. We conclude that the Time 2 reports provided a reasonably valid assessment for the presence (but not frequency) of CP behavior in the mothers' repertoire. A unique finding that must be considered preliminary because of the sample size is that the rate of CP observed far exceeds prior estimations or previously published finding. For example, American parents of 2-year-old children reported they spanked or slapped a mean of 18 times per year (Straus & Stewart, 1999). Here, mothers at the median frequency rate would be striking their children 18 times within one week\ Larzelere (2000) concluded that detrimental outcomes can occur when parents are spanking one to three times a week. Every single parent who used CP in this study was at or above that threshold. In the course of invesfigating customary CP in situ, most of the incidents were fleeting and unremarkable. However, coders were alarmed when hearing a mother hit her child 11 times in a row, the severe emotional distress reacfion from four children, and the 7-month-old infant who was hit. These incidents provide a peek into how a disciplinary incident could escalate whereby a young child could be inadvertently injured while being punished or a repeatedly distressed child could be emotionally affected. Although prospanking advocates may acknowledge these incidents as inappropriate uses of CP, evidence indicates that mothers who report their child gets spanked are also more likely to report physical abuse of that child (Zolotor, Theodore, Chang, Berkoff, & Runyan, 2008). Despite the novel results, both the audio-recording method and the study have limitations. The audio method was limited by self-selection into the study; although it was not assessed, many families may find audio recordings of home interactions far too intrusive of their privacy to give consent. Additionally, families who did not adhere to instructions and failed to record entire evenings of interactions limited the scope of audio recordings. There is also the obvious limitation that the recordings do not capture nonverbal information, so we could not observe where the children were hit nor their physical reactions. Finally, the audiorecording approach is extremely time intensive, both for data collection and coding. Our investigation into naturally occurring CP also has some notable limitations. Foremost, the small, selective sample of families contributed differing numbers of CP incidents, based on a differing number of hours of recordings. We also do not know the extent to which parents were reactive to the method and modified their typical behavior (Gardner, 2000). However, given that the CP
405
incidents were recorded as soon as the very first night, we do not think reactivity was a serious problem. In this report, we do not compare CP with other disciplinary practices; such a comparison is in progress. Because of the limitations, the results of this study should be regarded as preliminary. Nevertheless, advocates of CP as a disciplinary technique believe that if used appropriately, it is an effective technique. What these data indicate is that it is not effective in terminating misbehavior over as short a time period as 10 minutes. Moreover, "customary" CP does not appear to be used "prudently" (Bautnrind, 1997) as it falls far short of conforming to the recommended guidelines. The study illustrates that direct observations of the family, through the use of audio recordings, is a feasible method and can provide novel data. As this pilot effort illustrates, the method can shed light on important family interactions, such as whether parental use of CP is maligned by opponents or "misused" by parents in the home.
References Ateah, C. A., & Durrant, J. E. (2005). Maternal use of physical punishment in response to child misbehavior: Implications for child abuse prevention. Child Abuse & Neglect, 29, 169-185. doi:10.l016/j.chiabu.2004 .10.010, Baumeister, R. F., Vohs, K. D., & Funder, D. C. (2007). Psychology as the science of self-reports and finger movements: Whatever happened to actual behavior? Perspectives on Psychological Science, 2, 396-403. doi:10.1111/j.l745-6916.2007.0005I.x Baumrind, D. (1997). Necessary distinctions. Psychological Inquiry, 8, 176-182. doi: 10.1207/s 15327965pli0803_2 Baumrind, D. (2012). Differentiating between confrontive and coercive kinds of parental power-assertive disciplinary practices. Human Development, 55, 35-51. doi: 10.1159/000337962 Benjet, C . & Kazdin, A. E. (2003). Spanking children: The controversies, findings, and new directions. Clinical Psychology Review, 23, 197-224. doi: 10.1016/S0272-7358(02)00206-4 Bradley, R. H., Corwyn, R. F., McAdoo, H. P., & Garcia Coll, C. (2001). The home environments of children in the United States, Part I: Variations by age. ethnicity, and poverty status. Child Development, 72, 1844-1867. doi:10.1111/1467-8624.t01-l-00382 Brown. B. W. (1979). Parents' discipline of children in public places. The Family Coordinator, 28, 67-71. doi: 10.2307/583270 Davis, P. W. (1996). Threats of corporal punishment as verbal aggression: A naturalistic study. Child Abuse & Neglect, 20, 289-304. dol:10.1016/ 0145-2134(96)00010-5 Dobson, J. (1992). The new dare to discipline. Wheaton, IL: Tinsdale House. Dunn, J., & Munn, P. (1987). Development of justification in disputes with mother and sibling. Developmental Psychology, 23, 791-798. doi: 10.1037/0012-1649.23.6.791 Durrant, J., & Ensom, R. (2012). Physical punishment of children: Lessons from 20 years of research. Canadian Medical Association Journal. doi:10.1503/cmaj.101314 Gardner, F. (2000). Methodological issues in the direct observation of parent-child interaction: Do observational findings reflect the natural behavior of participants? Clinical Child and Family Psychology Review, 3, 185-198. doi:l0.1023/A:1009503409699 Gershoff, E. T. (2013). Spanking and child development: We know enough now to stop hitting our children. Child Development Perspectives, 7, 133-137. doi:10.1111/cdep.l2038 Gershoff, E. T., Grogan-Kaylor, A., Lansford, J. E., Chang, L., Zelli, A., Deater-Deckard, K., & Dodge, K. A. (2010). Parent discipline practices in an international sample: Associations with child behaviors and mod-
406
HOLDEN, WILLIAMSON, AND HOLLAND
eration by perceived normativeness, Chitd Development, 81, 487-502. doi:10.1111/j.l467-8624,2009.01409.x Holden, G. W,, Coleman, S., & Schmidt, K. L, (1995), Why 3-year-o!d children get spanked: Parent and child determinants in a sample of college-educated mothers, Merrill-Palmer Quarterly, 41, 431-452, Hops, H., Davis, B., & Longoria, N, (1995). Methodological issues in direct observation: Illustrations with the Living in Familial Environments (LIFE) coding system. Joumal of Clinical Child Psychology, 24, 193-203. doi: 10.1207/s 15374424jcep2402_7 Johnson, S. M., & Bolstad, 0. D. (1975), Reactivity to home observation: A comparison of audio recorded behavior with observers present or absent. Journal of Applied Behavioral Analysis, 8, 181-185, doi: 10.1901/jaba.l975,8-181 Lansford, J, E., Wager, L. B., Bates, J. E., Dodge, K, A,, & Pettit, G. S, (2012), Parental reasoning, denying privileges, yelling, and spanking: Ethnic differences and associations with child externalizing behavior. Parenting: Science and Practice, 12, 42-56. doi: 10,1080/15295192 ,2011.613727 Larzelere, R, E, (2000), Child outcomes of nonabusive and customary physical punishment by parents: An updated literature review. Clinical Child and Family Psychology Review, 3, 199-221, doi:10,1023/A: 1026473020315 Maguire-Jack, K., Gromoske, A, N,, & Berger, L, M, (2012). Spanking and child development during the first 5 years of life. Child Development, 83, 1960-1977, doi:10,l 11 l/j.l467-8624.2012,01820,x Matjasko, J. L., & Feldman, A, F, (2006). Bringing work home: The emotional experiences of mothers and fathers. Journal of Family Psychology, 20, 47-55. doi:10.!037/0893-3200.20.1.47 Morsbach, S. K., & Prinz, R, J, (2006), Understanding and improving the validity of self-report of parenting. Clinical Chitd and Family Psychology Review, 9, l-21,doi:10,1007/sl0567-006-0001-5
Robinson, C. C , Mandleco, B., Olsen, S, F„ & Hart, C, H. (1995). Authoritative, authoritarian, and permissive parenting practices: Development of a new measure. Psychological Reports, 77, 819-830. doi: 10.2466/prO.1995,77,3,819 Rosemond, J, (1994), To spank or not to spank: A parents' handbook. Kansas City, MO: Andrews & McMeel. Shiffman, S,, Stone, A. A., & Hufford, M. R. (2008). Ecological momentary assessment. Annual Review of Clinical Psychology, 4, 1-32, doi: 10,1146/annurev.clinpsy.3.022806.091415 Stansbury, K,, Haley, D,, Lee, J, A,, & Brophy-Herb, H. E. (2012). Adult earegivers' behavioral responses to child noncompliance in public settings: Gender differences and the role of positive and negative touch. Behavior and Social Issues, 21, 80-114, doi:10,5210/bsi,v21iO,2979 Su-aus, M. A., & Stewart, J. H. (1999). Corporal punishment by American parents: National data on prevalence, chronicity, severity, and duration in relation to child and family characteristics. Clinical Child and Family Psychology Review, 2, 55-70. doi:10.1023/A:1021891529770 Trumbull, D., & Ravenel, S. (1998). How to spank. Focus on the Family, 2-4. Vittrup, B., Holden, G. W,, & Buck, J, (2006), Attitudes predict the use of physical punishment: A prospective study of the emergence of disciplinary practices. Pediatrics, 117, 2055-2064. doi:10.1542/peds.2005-2204 Zolotor, A, J,, Theodore, A. D., Chang, J, J., Berkoff, M. C , & Runyan, D, K. (2008). Speak softly—And forget the stiek: Corporal punishment and child physical abuse. American Joumal of Preventive Medicine, 35, 364-369. doi:10.1016/j.amepre,2008,06,03!
Received July 29, 2013 Revision received February 6, 2014 Accepted February 11, 2014 •
Copyright of Journal of Family Psychology is the property of American Psychological Association and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
© 2014 American Psychological Association 0893-3200/14/S12.00 DOI: 10.1037/a0036370
BRIEF REPORT
Eavesdropping on the Family: A Pilot Investigation of Corporal Punishment in the Home George W. Holden, Paul A. Williamson, and Grant W. O. Holland Southern Methodist University This study tested the feasibility of using audio recorders to collect novel information about family interactions. Research into corporal punishment (CP) has relied, almost exclusively, on self-report data; audio recordings have the promise of revealing new insights into the use and immediate consequences of CP, So we eould hear how parents respond to child conflicts, 33 mothers wore digital audio recorders for up to 6 evenings. We identified a total of 41 CP incidents, in 15 families and involving 22 parent-child dyads. These incidents were evaluated on 6 guidelines culled from the writings of CP advocates. The results indicated, contrary to advice, CP was not being used in line with 3 of the 6 recommendations and for 2 others, the results were equivocal. The last recommendation could not be assessed with audio. Latency analyses revealed children, after being hit, were misbehaving again within 10 minutes after 73% of the incidents. Mothers' self reports about whether they used CP were found to correspond to the audio data in 81% of the cases. Among the mothers who were hitting, CP occurred at a much higher rate than the literature indicates. These results should be viewed as preliminary because of the small sample of families and the even smaller number of families who used CP. Nevertheless, this pilot study demonstrates that audio recording naturally occurring momentary processes in the family is a viable method for collecting new data to address important questions about family interactions. Keywords: audio recordings, corporal punishment, discipline, ecological momentary assessment, home observations
The overreliance on behavioral self-reports is a well-recognized limitation of contemporary research (e,g,, Baumeister, Vohs, & Funder, 2007), Self-reports are limited by recall abilities, lack of awareness, social desirability, and other sources of distortion (Morsbach & Prinz, 2006), Self-reports also limit the types of questions that can be addressed, such as with regard to sequential analyses or latencies. Observational data, though it has problems of its own (e,g,, Gardner, 2000; Hops, Davis, Longoria, 1995), is able to capture the sequence of actual ongoing behavior. In this study, we report on the feasibility of using audio recordings to provide new insights into one way in which parents respond to child misbehavior. Although audio recording family interactions is not a new methodology (e,g„ Johnson & Bolstad, 1975), it is rarely used. The advantage of audio recorders over video recorders is that because a camera operator is not needed, audio devices use
fewer resources and lessen the likelihood of reactivity (Gardner, 2000), Despite the absence of visual data, audio provides rich interaction data and can ufilize the ecological momentary assessment approach, characterized by repeated sampling of subjects in real fime in their natural environments (Shiffman, Stone, & Hufford, 2008). We used audio recordings to investigate how parents use corporal punishment (CP). This is a prime topic to study because the extensive literature on the topic relies almost exclusively on selfreports. This reliance on one method of data collection limits our understanding of CP and its impact. For example, parents may distort their reports (consciously or otherwise), they may be unaware of their own behavior, and they are likely unable to report, with fidelity, parent-child interaction sequences. It is also an important topic to investigate given the intemational debate over the use of corporal punishment (e,g,, Benjet & Kazdin, 2003; Durrant & Ensom, 2012), Despite its widespread use both in the United States and internationally (Gershoff et al, 2010; Lansford, Wager, Bates, Dodge, &. Pettit, 2012), opponents of CP argue that spanking or slapping is ineffective, and cite the large number of studies that link it to a variety of unintended child behavior problems and other negative sequelae (e,g„ Durrant & Ensom, 2012; Gershoff, 2013; MaguireJack, Gromoske, & Berger, 2012), On the other hand, proponents of CP are critical of the corpus of research, arguing that the data are correlational and have various methodological limitations, such as not controlling for child effects (e,g,, Larzelere, 2000), Advo-
This article was published Online First April 14, 2014, George W. Holden, Paul A. Williamson, and Grant W. O, Holland, Department of Psychology, Soudiem Methodist University. We thank the numerous research assistants who aided in the collection and coding of data. This research was supported by a grant from the Timberlawn Psychiatric Research Foundation. Correspondence concerning this article should be addressed to George Holden, Department of Psychology, Southern Methodist University, P.O. Box 750442, Dallas, TX 75275-0442. E-mail: gholden@smu,edu 401
HOLDEN, WILLIAMSON, AND HOLLAND
402
cates also point out that some studies "lump together" all types of corporal punishment, thereby conflating the effects of abusive practices with "normative" spanking or "customary" spanking. They hold that the effects of spanking depend on how it is used, and if used appropriately, it can be an effective form of power assertion (Baumrind, 1997, 2012). Audio recordings of family conflicts can help to fill in a major gap in the CP literature—that of the immediate and concomitant consequences of CP (Benjet & Kazdin, 2003). Despite the hundreds of studies investigating the associations between CP and child behavior, there continues to be a paucity of information about what transpires before, during, and immediately after disciplinary incidents involving CP. As a result, it is impossible to evaluate the validity of proponents' assertions that a) there is a "customary" form of CP, b) that customary CP follows suggested guidehnes, and c) that this form of CP is effective for disciplining children. Assessing the validity of these assertions requires naturalistic observations. To date, observations of parents engaging in CP are extremely rare. Brown (1979) observed a total of 20 "hitting or physical force" responses in a shopping mall. In a study of verbal aggression, Davis (1996) reported that he observed parents hitting their children in public, backing up threats of hitting about one third of the time. The Home Observation for the Measurement of the Environment (e.g., Bradley, Corwyn, McAdoo, & Garcia Coll, 2001) measures CP simply as present or absent while the interviewer is present. Recently, Stansbury, Haley, Lee and BrophyHerb (2012) observed "negative touches" (slapping, pinching, hitting, or threatening to hit) in response to noncompliance in public settings. In 23% of the incidents, children received negative touches. The reports of these observational studies do not provide enough information to determine whether the parents followed prescribed recommendations or whether the CP was effective in the immediate context. Nor do they shed light on the proximal determinants of CP, such as the eliciting misdeed or parental affect (Ateah & Durrant, 2005). In the present study, we use home observations to examine the use of CP and to assess its immediate impact on children's behavior. We evaluate how it is used in light of "best practices" guidelines provided by proponents. Seven guidelines were identified; a) it should be used infrequently; b) it should be used selectively, for serious misbehaviors, such as aggression; c) it should be used as a last resort; d) it should be administered calmly, not in anger; e) it should consist of no more than two hits; f) the child should be struck on the buttocks; and f) it should be painful. These guidelines, gleaned by Vittmp, Holden, and Buck (2006), came from five sources (e.g., Dobson, 1992; Rosemond, 1994; Trumbull & Ravenel, 1998). In sum, we sought to explore whether home audio recordings could reveal new information about CP use in the home and whether it adhered to the parameters recommended by proponents of the disciplinary technique. We also investigated the underlying assumption of proponents that CP is effective at curtailing subsequent misbehavior. Einally, we addressed a methodological question: Do self-reports accurately reflect the observed frequency of CP?
Method The study complied with the APA's ethical principles and was approved by the university's IRB committee.
Participants Participants were recmited using flyers distributed to Englishspeaking mothers of 2- to 5-year-old children at daycare and Head Start centers in a large southwestem city. A total of 56 potential participants contacted the investigators. Only the mothers who reported, in a phone screen about disciplinary practices, that they yelled in anger at least twice a week (the original study focus) were invited to participate. Eourteen mothers did not meet this criterion. In addition, seven mothers dropped out. Attrition analyses on the screening data showed the drop-outs did not differ (race/ethnicity, number of children, or frequency of spanking) from the participants. Thirty-five mothers of 2- to 5-year-old children participated. Audio recordings from two families were insufficiently intelligible to use. In the final sample of 33 families, the target children averaged 46.0 months of age (SD = 11.8, range 24 to 68); 13 were female. The median number of siblings was one (range 0 to 3). The mothers averaged 33.8 years old (SD = 5.8, range 23 to 46), most were married (82.1 %), and self-reported as white (60.7%), African American (25.0%), or Hispanic (10.7%). The mothers were employed full-time outside the home (60.7%), worked part-time (17.9%), were homemakers (14.3%), or were students (3.6%). One third of the mothers (32.1%) had attended some college or vocational school, another third (32.1%) were college graduates, and 28.6% had advanced degrees or training. Eathers' average age was 35.7 years (SD = 4.5) and most (91%) were employed full time. Their education ranged from high school (14%) to graduate or professional school (24%). Annual family income included less than $40,000 (28.6%), between $40,000 and $79,999 (32.2%), $80,000 or more (39.2%).
Measures Audio recordings. Mothers, visited at their homes, were given an Olympus DS-50 digital voice recorder to wear, in a sport pouch, on the upper arm. They were instmcted to turn on the recorder upon returning home from work (or at 5:00 p.m.) and turn it off once their child fell asleep. The first 10 participants were asked to make recordings on four consecutive days; the remaining participants were asked to record six consecutive days. This procedure resulted in an average of 12.95 hrs of recordings per family (SD = 6.16, range 2.1 to 25.5). Most (94%) of the mothers complied with the number of recordings requested, though only 62% of the recordings continued all the way through bedtime. When the recordings were completed, a researcher retumed to the home to collect the audio recorder and questionnaires. Mothers received $120 for their efforts. The first step in coding required identifying the CP incidents. In 51% of the incidents, the sound of the child being slapped or spanked was clearly discernible and supported by contextual cues, such as wamings of or justifications for the hit. In 44% of the incidents, the sound of CP was ambiguous, but the contextual cues (mother's warning of impending CP, child's cries) provided supporting evidence. In two cases (5%), there was no audible sound of CP but clear explicit contextual information, such as the child pleading, "Stop hitting me." Two raters independently coded 30% of the recordings. The interrater reliability, computed by agreements over agreements and disagreements, was .87, with a Co-
EAVESDROPPING ON THE FAMILY hen's kappa (K) of .77. Disagreements for all codes were discussed and settled by consensus. The second step involved a detailed analysis by two coders of each incident relating to the seven guidelines. However, given that we were relying on audio data, the location on the body as to where the child was struck could not be determined. Therefore, that guideline was not evaluated. We derived the dependent variables for the other six guidelines in the following ways: CP should be used infrequently. To measure the frequency of CP use, we calculated the rate of CP per hour of recording. Because only mothers wore the recorders, this analysis was limited to them. The interrater reliability was .87, K = .77. CP should be used selectively, for serious behaviors. To assess the types of behaviors eliciting CP, we coded the child transgressions into five categories: a) social convention (e.g., personal untidiness or uncleanliness, poor manners); b) aggression (verbal or physical acts that could hurt someone); c) rights violation (e.g., taking someone's property; intruding on other's personal space); d) destruction (intentional acts that could damage objects); e) prudential rule violation (engaging in a safety risk such as touching a stove). The first four categories came from Dunn and Munn (1987). All but two of the CP incidents could be coded into one of these categories. The reliability was .73, K = .58. CP should be used as a last resort. This guideline was operationalized in two ways. First, we measured the latency in seconds from the onset of the child's infraction until the CP. The intraclass correlation between the two raters was .71. Second, we summed the number of responses to the transgression prior to the first hit. The agreement was .86; ICC = .44. CP should not be used in anger. We assessed, on a dichotomous code, the presence of parental negative affect when the CP was delivered. Negative affect occurred if the volume of the utterance increased, the tone of voice indicated aggravation or anger, and/or the presence of negative verbalizations (e.g., threats). The interrater reliability was M = .73, K = .46. The CP should consist of no more than two hits. Based on the sound or verbal comments, we counted the number of slaps/ spanks. The reliability was very high, ICC = .99. The hit should be painful. We operationalized the degree of pain experienced by the child on a 4-point distress scale (0 = no audible response; 1 = minimal but audible, such as whimpering; 2 = moderate, may include some crying or verbal protest; 3 = strong, such as loud crying). The interrater reliability was .71, K = .53. "Appropriate" CP is effective. Effectiveness of CP was operationalized as whether a new conflict occurred within 10 min of the CP. The agreement (within 5 s) was .90, K = .78.
Questionnaires To assess the correspondence between self-reported CP and observed frequency of CP, we administered two questionnaires. Each mother completed the PRCM twice: once at the time of screening (Time 1) and once at home during the period when recordings were being made (Time 2). The PSDQ was only filled out at Time 2. Parental Responses to Child Misbehavior (PRCM). The PRCM (Holden et al., 1995) asks parents how frequently, in an average week, they use particular disciplinary responses, rated on
403
7-point scales (0 = Never, 1 = less than once; 2 = 1-2 times; 3 = 3-4 times; 4 = 5-6 times; 5 = 7-8 times; 6 = 9 or more times). The original PRCM has good test-retest reliability, internal consistency, and validity (e.g.. Holden et al., 1995; Vittrup et al., 2006). The original PRCM contained only one item referring to CP ("spank/slap"). For the present study, we revised that item to "spank with hand" and added three more CP items: "spank with an object," "slap on the child's hand," and "slap on the child's face." However, the last item lowered the internal consistency of the new four-item CP subscale. Consequently we removed it. The final 3-item PRCM-CP subscale correlated strongly across Time 1 and Time 2 (r = .75, p < .001), but was only moderately internally consistent (a^ime i = -52, a.rime 2 = -61). Parenting Styles and Dimensions Questionnaire (PSDQ). The PSDQ (Robinson, Mandleco, Olsen, & Hart, 1995) is a commonly used 62-item measure. For the purposes of this study we report only on the six-item CP subscale (e.g., "I spank when our child is disobedient"), rated on a 5-point frequency scale (1 = Never; 5 = Always), with an a of .70 in this sample.
Results CP Incidents A total of 41 CP incidents were heard in 15 of the 33 families (45% of the sample). The distribution of incidents within these families ranged widely. Six families each contributed one incident, four families had two incidents, one family had three incidents, two families had four incidents, one family had six incidents, and the last family had 10 incidents. Only mothers used CP in nine families, only fathers in three families, both parents in two families, and mother and grandmother in one family. (For convenience, we will refer to all these individuals as "parents.") Eighteen children (11 boys) received CP. Their median age was 48 months (range 7 months to 6 years). Twelve mothers accounted for 32 incidents, five fathers for seven incidents, and one grandmother for two incidents. Because some families contained multiple children and two adults, we calculated the total number of parent-child dyads involved. The incidents involved 22 different parent-child dyads. Sixteen of the parent-child dyads had one incident, one dyad experienced two incidents, two dyads had three, and one dyad each contributed four, five, and 10 incidents. Three children were hit by both their parents, and three parents hit two children. The incidents were more frequent in the first three evenings (six, 11, & 10 incidents, respectively) than in the last three (six, one, & seven, respectively).
Evaluating CP Incidents in Light of Proponents' Guidelines Next we summarize the correspondence between the audio data and the six guidelines. CP should be used infrequently. The median rate was .22/hr (SD = .16) with a low of .06 per hr (based on one incident for 17.95 hrs) to a high of 1.43/hr (three incidents from 2.1 hrs). CP should be used selectively, for serious behaviors. To evaluate selectivity in use, we examined the eliciting misbehavior. In 40 of the 41 incidents we identified the misdeed or source of the
404
HOLDEN, WILLIAMSON, AND HOLLAND
confiict. Noncompliance was the proximate cause in 90% of the incidents. Typically, this noncompliance consisted of the child ignoring or resisting parental instruction or command ("Come here!"). The immediate cause of the remaining four incidents was aggression or breaking a standing rule (resulting in CP without warning). The misdeed that precipitated the noncompliance was coded into the five behavioral domains. Seventy-four percent of the transgressions involved breaking social conventions. The offending misbehaviors consisted of such actions as sucking fingers, eating improperly, getting out of a chair, and going outside without permission. The next most frequent category (10%) was acts of aggression, including the child hitting someone or throwing an object at the parent. Rights-related misdeeds occurred in 7.5% of the incidents. The remaining misdeeds were prudential (5%) or destructive (2.5%). CP should be used as a last resort. On average, parents tried one disciplinary response before the CP (M = 1 . 1 , SD = .97, range 0 to 3). Most commonly, the initial reaction was a prohibition (e.g., "Stop it!"). The elapsed time from the onset of a confiict to administration of the CP averaged less than 30 s (M = 26.5, SD = 67.31, range 0 to 416). CP should not be used in anger. In 49% of the incidents, parents expressed negative affect while delivering the CP. The CP should consist of no more than two hits. Only one hit was audible in 83% of the incidents. In three cases (7.3%), two hits were heard. There was one incident each with 3, 5, 6, and 11 hits. The mean number of hits was 1.59 {SD = 1.83). The hit should be painful. The modal child distress rating was moderate (48.8%), followed by minimal (29.3%), and strong negative reaction (9.8%). No audible child reaction was heard in 12.2% of the incidents. The mean distress rating was 1.56 {SD = .84). Finally, to gauge the effectiveness of CP, we coded whether the cMld misbehaved during the immediate 10 tnin that followed the CP. In 30 of the 41 CP incidents (73%), the children engaged in the same or another misbehavior within that time period. To supplement these descriptive statistics, we conducted multilevel analyses in an effort to account for nonindependence of observations. Those analyses did not differ substantially from the results presented here, so are not reported (available from the first author).
Correspondence Between the Audio and Questionnaire Data We examined the correspondence between the two forms of data in two ways. The correlation between the CP subscale on the PRCM (Tl) and the actual number of CP incidents was not significant, r = .21, p = .23. However, when using the CP subscale from the Time 2 (r[30] = .76, p < .001 with Tl), the correlation increased to r = .54, p < .001. That can be accounted for by the tendency for mothers to report more CP at Time 2{M = 2.87) than at Time 1 (M = 2.33), paired r(29) = 1.44, p = .16. The CP subscale from the PSDQ survey also correlated significantly with incidents, r = .47, p < .001, as well as with the PRCM subscale (r^-j^^ , = .58, r.^^^^ ^ = -68, ps < .001). When examining reports of CP dichotomously (spank vs. no spank) based on whether mothers reported they used CP at least 1-2 times per week
on the PRCM (T2), reports and behavior showed good correspondence, x^(l) = 11.79, p = .001. Seventeen mothers reported they did not use CP (or did so less than once/week) and were not heard to use CP. Nine mothers who reported they did use CP, indeed used it. Four mothers reported they did use CP but were not heard to; two mothers reported they did not use CP but used it.
Discussion The use of audio recorders proved to be a highly effective technique to study the moment-to-moment interactions surrounding conflict in the home. The method afforded new data about how "customary" CP was being used in the home—at least by this small sample of parents. The audio recordings allowed us to assess CP use in the home and examine several new variables related to the CP use that had not been previously reported in the literature. The recordings revealed CP occurred, on average, just 30 s after the onset of a conflict and with little warning. These conflicts typically resulted from the violation of a social convention. Parents, clearly angry in half the cases, typically hit their children once or twice per incident. Within 10 min of the physical discipline, in 73% of the events examined, children misbehaved again. The recordings revealed that CP was frequently not being used in the ways that advocates of the disciplinary technique recommended. Three caveats about generalizability are in order. First, this small self-selected sample of parents was screened for yelling and the mothers are more educated than the general population. It is possible these parents used more frequent, harsher disciplinary practices than would be found in a more representative sample. In addition, the recordings were made during the most stressful time period of the day, when emotional spillover from prior events are likely and children are most at-risk for CP (Holden et al, 1995; Matjasko & Feldman, 2006). Therefore, the observed rates of CP would likely be lower if recordings had been made at other times of the day. Third, CP was heard in only 22 parent-child dyads from 15 families. Evidently, this study should be considered a pilot effort and a larger, more representative sample is needed. Of the parents who used CP, their behavior commonly violated three of the six "use" guidelines examined; infrequent, serious misbehavior, and last resort. Among the 12 mothers who engaged in CP, the median rate was once every 6.3 hrs of interactions. In about three-quarters of the incidents, parents hit their children for extraordinarily mundane social convention offenses, rather than serious infractions, such as rights or prudential violations (Dunn & Munn, 1987). Most poignantly, one child was slapped for turning pages of a storybook. Parents used CP as their second, not last, resort, on average just half a minute after the conflict began. With only a few exceptions, parents were abiding with one of the guidelines—hitting only once or twice. The data were equivocal for the last two guidelines. Although the guidelines call for calmly metering out punishment, parents were rated as angry in about half of the incidents. Consequently, these data suggest that the role of negative emotion, while present in many incidents, should not be exaggerated as a determinant (Ateah & Durrant, 2005). The final guideline (making CP painful) was indirectly assessed with a rating how upset the child sounded. In just over half of the cases, the children were heard to be at least moderately
EAVESDROPPING ON THE FAMILY
upset after the CP. Videos would provide better assessments of children's emotional response. When all six guidelines are considered together in one incident, it is clear that parents were rarely, if ever, using CP "appropriately." The recordings reveal that most parents were responding either impulsively and/or emotionally, rather than instrumentally and intentionally (Benjet & Kazdin, 2003). Impulsivity may have attenuated the correlation between self-reports of CP and observed incidents. Nevertheless, at Time 2, when the mothers were somewhat more forthcoming about their CP behavior, the x^ test was significant for the reports of and overheard CP incidents, matching in 81% of the cases. Most of the disagreements (12%) involved mothers who reported they used CP but were not heard to. This discrepancy is not surprising given the recordings averaged about 13 hrs over four to six days and mothers were reporting on behavior across an average week. We conclude that the Time 2 reports provided a reasonably valid assessment for the presence (but not frequency) of CP behavior in the mothers' repertoire. A unique finding that must be considered preliminary because of the sample size is that the rate of CP observed far exceeds prior estimations or previously published finding. For example, American parents of 2-year-old children reported they spanked or slapped a mean of 18 times per year (Straus & Stewart, 1999). Here, mothers at the median frequency rate would be striking their children 18 times within one week\ Larzelere (2000) concluded that detrimental outcomes can occur when parents are spanking one to three times a week. Every single parent who used CP in this study was at or above that threshold. In the course of invesfigating customary CP in situ, most of the incidents were fleeting and unremarkable. However, coders were alarmed when hearing a mother hit her child 11 times in a row, the severe emotional distress reacfion from four children, and the 7-month-old infant who was hit. These incidents provide a peek into how a disciplinary incident could escalate whereby a young child could be inadvertently injured while being punished or a repeatedly distressed child could be emotionally affected. Although prospanking advocates may acknowledge these incidents as inappropriate uses of CP, evidence indicates that mothers who report their child gets spanked are also more likely to report physical abuse of that child (Zolotor, Theodore, Chang, Berkoff, & Runyan, 2008). Despite the novel results, both the audio-recording method and the study have limitations. The audio method was limited by self-selection into the study; although it was not assessed, many families may find audio recordings of home interactions far too intrusive of their privacy to give consent. Additionally, families who did not adhere to instructions and failed to record entire evenings of interactions limited the scope of audio recordings. There is also the obvious limitation that the recordings do not capture nonverbal information, so we could not observe where the children were hit nor their physical reactions. Finally, the audiorecording approach is extremely time intensive, both for data collection and coding. Our investigation into naturally occurring CP also has some notable limitations. Foremost, the small, selective sample of families contributed differing numbers of CP incidents, based on a differing number of hours of recordings. We also do not know the extent to which parents were reactive to the method and modified their typical behavior (Gardner, 2000). However, given that the CP
405
incidents were recorded as soon as the very first night, we do not think reactivity was a serious problem. In this report, we do not compare CP with other disciplinary practices; such a comparison is in progress. Because of the limitations, the results of this study should be regarded as preliminary. Nevertheless, advocates of CP as a disciplinary technique believe that if used appropriately, it is an effective technique. What these data indicate is that it is not effective in terminating misbehavior over as short a time period as 10 minutes. Moreover, "customary" CP does not appear to be used "prudently" (Bautnrind, 1997) as it falls far short of conforming to the recommended guidelines. The study illustrates that direct observations of the family, through the use of audio recordings, is a feasible method and can provide novel data. As this pilot effort illustrates, the method can shed light on important family interactions, such as whether parental use of CP is maligned by opponents or "misused" by parents in the home.
References Ateah, C. A., & Durrant, J. E. (2005). Maternal use of physical punishment in response to child misbehavior: Implications for child abuse prevention. Child Abuse & Neglect, 29, 169-185. doi:10.l016/j.chiabu.2004 .10.010, Baumeister, R. F., Vohs, K. D., & Funder, D. C. (2007). Psychology as the science of self-reports and finger movements: Whatever happened to actual behavior? Perspectives on Psychological Science, 2, 396-403. doi:10.1111/j.l745-6916.2007.0005I.x Baumrind, D. (1997). Necessary distinctions. Psychological Inquiry, 8, 176-182. doi: 10.1207/s 15327965pli0803_2 Baumrind, D. (2012). Differentiating between confrontive and coercive kinds of parental power-assertive disciplinary practices. Human Development, 55, 35-51. doi: 10.1159/000337962 Benjet, C . & Kazdin, A. E. (2003). Spanking children: The controversies, findings, and new directions. Clinical Psychology Review, 23, 197-224. doi: 10.1016/S0272-7358(02)00206-4 Bradley, R. H., Corwyn, R. F., McAdoo, H. P., & Garcia Coll, C. (2001). The home environments of children in the United States, Part I: Variations by age. ethnicity, and poverty status. Child Development, 72, 1844-1867. doi:10.1111/1467-8624.t01-l-00382 Brown. B. W. (1979). Parents' discipline of children in public places. The Family Coordinator, 28, 67-71. doi: 10.2307/583270 Davis, P. W. (1996). Threats of corporal punishment as verbal aggression: A naturalistic study. Child Abuse & Neglect, 20, 289-304. dol:10.1016/ 0145-2134(96)00010-5 Dobson, J. (1992). The new dare to discipline. Wheaton, IL: Tinsdale House. Dunn, J., & Munn, P. (1987). Development of justification in disputes with mother and sibling. Developmental Psychology, 23, 791-798. doi: 10.1037/0012-1649.23.6.791 Durrant, J., & Ensom, R. (2012). Physical punishment of children: Lessons from 20 years of research. Canadian Medical Association Journal. doi:10.1503/cmaj.101314 Gardner, F. (2000). Methodological issues in the direct observation of parent-child interaction: Do observational findings reflect the natural behavior of participants? Clinical Child and Family Psychology Review, 3, 185-198. doi:l0.1023/A:1009503409699 Gershoff, E. T. (2013). Spanking and child development: We know enough now to stop hitting our children. Child Development Perspectives, 7, 133-137. doi:10.1111/cdep.l2038 Gershoff, E. T., Grogan-Kaylor, A., Lansford, J. E., Chang, L., Zelli, A., Deater-Deckard, K., & Dodge, K. A. (2010). Parent discipline practices in an international sample: Associations with child behaviors and mod-
406
HOLDEN, WILLIAMSON, AND HOLLAND
eration by perceived normativeness, Chitd Development, 81, 487-502. doi:10.1111/j.l467-8624,2009.01409.x Holden, G. W,, Coleman, S., & Schmidt, K. L, (1995), Why 3-year-o!d children get spanked: Parent and child determinants in a sample of college-educated mothers, Merrill-Palmer Quarterly, 41, 431-452, Hops, H., Davis, B., & Longoria, N, (1995). Methodological issues in direct observation: Illustrations with the Living in Familial Environments (LIFE) coding system. Joumal of Clinical Child Psychology, 24, 193-203. doi: 10.1207/s 15374424jcep2402_7 Johnson, S. M., & Bolstad, 0. D. (1975), Reactivity to home observation: A comparison of audio recorded behavior with observers present or absent. Journal of Applied Behavioral Analysis, 8, 181-185, doi: 10.1901/jaba.l975,8-181 Lansford, J, E., Wager, L. B., Bates, J. E., Dodge, K, A,, & Pettit, G. S, (2012), Parental reasoning, denying privileges, yelling, and spanking: Ethnic differences and associations with child externalizing behavior. Parenting: Science and Practice, 12, 42-56. doi: 10,1080/15295192 ,2011.613727 Larzelere, R, E, (2000), Child outcomes of nonabusive and customary physical punishment by parents: An updated literature review. Clinical Child and Family Psychology Review, 3, 199-221, doi:10,1023/A: 1026473020315 Maguire-Jack, K., Gromoske, A, N,, & Berger, L, M, (2012). Spanking and child development during the first 5 years of life. Child Development, 83, 1960-1977, doi:10,l 11 l/j.l467-8624.2012,01820,x Matjasko, J. L., & Feldman, A, F, (2006). Bringing work home: The emotional experiences of mothers and fathers. Journal of Family Psychology, 20, 47-55. doi:10.!037/0893-3200.20.1.47 Morsbach, S. K., & Prinz, R, J, (2006), Understanding and improving the validity of self-report of parenting. Clinical Chitd and Family Psychology Review, 9, l-21,doi:10,1007/sl0567-006-0001-5
Robinson, C. C , Mandleco, B., Olsen, S, F„ & Hart, C, H. (1995). Authoritative, authoritarian, and permissive parenting practices: Development of a new measure. Psychological Reports, 77, 819-830. doi: 10.2466/prO.1995,77,3,819 Rosemond, J, (1994), To spank or not to spank: A parents' handbook. Kansas City, MO: Andrews & McMeel. Shiffman, S,, Stone, A. A., & Hufford, M. R. (2008). Ecological momentary assessment. Annual Review of Clinical Psychology, 4, 1-32, doi: 10,1146/annurev.clinpsy.3.022806.091415 Stansbury, K,, Haley, D,, Lee, J, A,, & Brophy-Herb, H. E. (2012). Adult earegivers' behavioral responses to child noncompliance in public settings: Gender differences and the role of positive and negative touch. Behavior and Social Issues, 21, 80-114, doi:10,5210/bsi,v21iO,2979 Su-aus, M. A., & Stewart, J. H. (1999). Corporal punishment by American parents: National data on prevalence, chronicity, severity, and duration in relation to child and family characteristics. Clinical Child and Family Psychology Review, 2, 55-70. doi:10.1023/A:1021891529770 Trumbull, D., & Ravenel, S. (1998). How to spank. Focus on the Family, 2-4. Vittrup, B., Holden, G. W,, & Buck, J, (2006), Attitudes predict the use of physical punishment: A prospective study of the emergence of disciplinary practices. Pediatrics, 117, 2055-2064. doi:10.1542/peds.2005-2204 Zolotor, A, J,, Theodore, A. D., Chang, J, J., Berkoff, M. C , & Runyan, D, K. (2008). Speak softly—And forget the stiek: Corporal punishment and child physical abuse. American Joumal of Preventive Medicine, 35, 364-369. doi:10.1016/j.amepre,2008,06,03!
Received July 29, 2013 Revision received February 6, 2014 Accepted February 11, 2014 •
Copyright of Journal of Family Psychology is the property of American Psychological Association and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
