PROOF COVER SHEET

Author(s): S.R. Wilson et al.Article title: Parental Self-efficacy and Sensitivity during Playtime Interactions

with Young Children: Unpacking the Curvilinear AssociationArticle no: RJAC 911937Enclosures: 1) Query sheet

2) Article proofs

Dear Author,1. Please check these proofs carefully. It is the responsibility of the correspondingauthor to check these and approve or amend them. A second proof is notnormally provided. Taylor & Francis cannot be held responsible for uncorrectederrors, even if introduced during the production process. Once your correctionshave been added to the article, it will be considered ready for publication.

Please limit changes at this stage to the correction of errors. You should not maketrivial changes, improve prose style, add new material, or delete existing materialat this stage. You may be charged if your corrections are excessive (we would notexpect corrections to exceed 30 changes).

For detailed guidance on how to check your proofs, please paste this address into anew browser window:http://journalauthors.tandf.co.uk/production/checkingproofs.asp

Your PDF proof file has been enabled so that you can comment on the proofdirectly using Adobe Acrobat. If you wish to do this, please save the file to yourhard disk first. For further information on marking corrections using Acrobat,please paste this address into a new browser window: http://journalauthors.tandf.co.uk/production/acrobat.asp

2. Please review the table of contributors below and confirm that the first andlast names are structured correctly and that the authors are listed in the correctorder of contribution. This check is to ensure that your name will appearcorrectly online and when the article is indexed.

Sequence Prefix Given name(s) Surname Suffix

1 Steven R. Wilson

2 Patricia E. Gettings

3 Lisa M. Guntzviller

4 Elizabeth A. Munz

Queries are marked in the margins of the proofs, and you can also click thehyperlinks below.

AUTHOR QUERIES

General points:

1. Permissions: You have warranted that you have secured the necessary writtenpermission from the appropriate copyright owner for the reproduction of anytext, illustration, or other material in your article. Please see http://journalauthors.tandf.co.uk/permissions/usingThirdPartyMaterial.asp.

2. Third-party content: If there is third-party content in your article, pleasecheck that the rightsholder details for re-use are shown correctly.

3. Affiliation: The corresponding author is responsible for ensuring thataddress and email details are correct for all the co-authors. Affiliationsgiven in the article should be the affiliation at the time the researchwas conducted. Please see http://journalauthors.tandf.co.uk/preparation/writing.asp.

4. Funding: Was your research for this article funded by a funding agency? Ifso, please insert ‘This work was supported by <insert the name of thefunding agency in full>’, followed by the grant number in square brackets‘[grant number xxxx]’.

5. Supplemental data and underlying research materials: Do you wish toinclude the location of the underlying research materials (e.g. data, samplesor models) for your article? If so, please insert this sentence before thereference section: ‘The underlying research materials for this article can beaccessed at <full link>/ description of location [author to complete]’. If yourarticle includes supplemental data, the link will also be provided in thisparagraph. See <http://journalauthors.tandf.co.uk/preparation/multimedia.asp> for further explanation of supplemental data and underlying researchmaterials.

6. The CrossRef database (www.crossref.org/) has been used to validate thereferences. Mismatches will have resulted in a query.

AQ1 As per journal style, reference citations are not allowed in the Abstract section.Hence, the reference citation [Coleman & Karraker, 2003], has been removed.Please check.

AQ2 Please provide the spell out form of “H2” at its first mention in the text.AQ3 Please spell out “GED” in full at its first mention.AQ4 Please spell out ‘IRB“ in full at its first mention, in the text.AQ5 Please note that we have introduced heading column as ”Excerpt 3 (LSM 4)“

and ”Excerpt 4 (LSM 33)” for Excerpt 3 and Excerpt 4, respectively, as per theformat followed for Excerpts 1 and 2 in the article. Please check and confirmwhether the same is ok.

AQ6 Please spell out “ICC” in full at its first mention, in the text.

AQ7 Has “Munz & Wilson, in press” been published yet? If so, please give details forreferences list.

AQ8 Webster-Stratton and Hammond, 1997, the pages have been replaced usingdata from CrossRef. Please provide the correct version if these are incorrect.

How to make correc�ons to your proofs using Adobe Acrobat/Reader

Taylor & Francis offers you a choice of op�ons to help you make correc�ons to your proofs.Your PDF proof file has been enabled so that you can edit the proof directly using Adobe Acrobat/Reader. This is the simplest and best way for you to ensure that your correc�ons will be

incorporated. If you wish to do this, please follow these instruc�ons:

1. Save the file to your hard disk. 2. Check which version of Adobe Acrobat/Reader you have on your computer. You can do this by clicking on the “Help” tab, and then “About”. If Adobe Reader is not installed, you can get the latest version free from h�p://get.adobe.com/reader/.

3. If you have Adobe Acrobat/Reader 10 or a later version, click on the “Comment” link at the right-hand side to view the Comments pane.

4. You can then select any text and mark it up for dele�on or replacement, or insert new text as needed. Please note that these will clearly be displayed in the Comments pane and secondary annota�on is not needed to draw a�en�on to your correc�ons. If you need to include new sec�ons of text, it is also possible to add a comment to the proofs. To do this, use the S�cky Note tool in the task bar. Please also see our FAQs here: h�p://journalauthors.tandf.co.uk/produc�on/index.asp.

5. Make sure that you save the file when you close the document before uploading it to CATS using the “Upload File” bu�on on the online correc�on form. If you have more than one file, please zip them together and then upload the zip file.

If you prefer, you can make your correc�ons using the CATS online correc�on form.

Troubleshoo�ng

Acrobat help: h�p://helpx.adobe.com/acrobat.htmlReader help: h�p://helpx.adobe.com/reader.html

Please note that full user guides for earlier versions of these programs are available from the Adobe Help pages by clicking on the link “Previous versions” under the “Help and tutorials” heading from the relevant link above. Commen�ng func�onality is available from Adobe Reader 8.0 onwards and from Adobe Acrobat 7.0 onwards.

Firefox users: Firefox’s inbuilt PDF Viewer is set to the default; please see the following for instruc�ons on how to use this and download the PDF to your hard drive: h�p://support.mozilla.org/en-US/kb/view-pdf-files-firefox-without-downloading-them#w_using-a-pdf-reader-plugin

Parental Self-efficacy and Sensitivityduring Playtime Interactions with YoungChildren: Unpacking the CurvilinearAssociation

5 Steven R. Wilson, Patricia E. Gettings, Lisa M. Guntzviller &Elizabeth A. Munz

This study shows that the association between parental self-efficacy (PSE) and observerratings of parents’ sensitivity during free play is curvilinear. Parents (33 mothers, 5

10 fathers) from an intergenerational learning program serving low-income familiescompleted the Self-Efficacy for Parenting Tasks Index AQ1and were videotaped during a15-minute play period with their infant/toddler at program entry. Parents’ self-reported PSE is positively, albeit weakly, associated with observer ratings of parentalsensitivity from low to moderate levels of self-efficacy but inversely associated with

15 sensitivity from moderate to high self-efficacy. Qualitative analyses show that motherswho report high PSE but are rated as low in sensitivity introduce toys when their childalready is engaged, restrict their child’s access to toys, physically manipulate their child,and violate their child’s proximal space. Parenting education programs need to developstrategies for reaching parents who enter with high levels of confidence but lack

20 knowledge of competent parenting practices.

Keywords: Parental Self-efficacy; Parental Sensitivity; Observational Methods; Parent–Child Interaction; Parenting Programs

Steven R. Wilson is a Professor and the Charles and Ann Redding Faculty Fellow in the Brian Lamb School ofCommunication, where Patricia E. Gettings is a doctoral student. Lisa M. Guntzviller is an Assistant Professor inthe Department of Languages, Philosophy, and Communication Studies at Utah State. Elizabeth A. Munz is anAssistant Professor in the Department of Communication Studies at West Chester University. Correspondence to:Steven R. Wilson, Brian Lamb School of Communication, Purdue University, West Lafayette, IN, USA. E-mail:wilson25@purdue.edu

Journal of Applied Communication ResearchVol. 00, No. 00, Month 2014, pp. 1–23

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:46 CE: M.V.; QA: N.D.

ISSN 0090-9882 (print)/ISSN 1479-5752 (online) © 2014 National Communication Associationhttp://dx.doi.org/10.1080/00909882.2014.911937

Parenting young children is rewarding but taxing as parents cope with sleepdeprivation, the need to constantly monitor their children’s safety, and the increasing

25 independence that emerges with toddlerhood while striving to communicate in waysthat promote their children’s healthy development. Parents vary in their knowledge,abilities, and confidence in carrying out the parenting role. Parenting self-efficacy(PSE) refers to “parents’ perceived ability to positively influence the behavior anddevelopment of their children” (Coleman & Karraker, 2003, p. 128). A large body of

30 literature links PSE to markers of parental competence such as sensitivity, consistency,and the use of nonpunitive discipline with children ranging from infancy toadolescence (Coleman & Karraker, 1997; Jones & Prinz, 2005). For this reason, PSEoften is used as one outcome for evaluating parenting programs.

Although evidence clearly supports the importance of PSE, there are hints in the35 literature that parents of young children sometimes report high levels of PSE yet fail

to perform behaviors associated with parenting competence or sensitivity (Jones &Prinz, 2005). Observational methods are essential for assessing sensitivity since itexamines the degree to which parents respond quickly and appropriately to theirchild’s social signals or gestures (Ainsworth, 1967). Some parents may hold idealized

40 views of their PSE, which in turn may result in the relationship between PSE andparenting sensitivity not being strictly linear. Understanding this potential complex-ity has implications for parenting programs since a subset of parents may enroll insuch programs with high but unrealistic confidence in their parenting knowledge andabilities.

45 This study examines the degree of sensitivity parents with varying levels of PSEdisplay while playing with their infants or toddlers during an unstructured playperiod. Data were gathered with low-income families enrolling in an intergenera-tional learning program in the authors’ local community. Parents, at program entry,completed a measure of PSE and were rated, based on their verbal and nonverbal

50 behaviors, for the degree of sensitivity they displayed while engaged in free play. Wealso conduct in-depth qualitative analyses of a subset (n = 8) of playtime interactionsto gain more insight about mothers who report high PSE but are rated as low insensitivity. The discovery that associations between PSE and parenting competencemay be more complicated than previously thought would have implications for

55 theories of parent–child interaction as well as intergenerational learning programs.

PSE and Parenting Competence

Self-efficacy involves a person’s judgment about his or her own ability to complete agiven task or action effectively in a specified setting. Bandura (1982, 1986, 1989)argues that self-efficacy is pivotal to how we interact with the environment and serves

60 as a central mediator between knowledge and behavior. Consistent with Bandura’s(1986) view of self-efficacy as domain specific, researchers have turned attention toself-efficacy regarding the parental role. PSE refers to the expectations caregivers holdabout their ability to be successful parents. According to Bandura’s framework, PSE

2 S.R. Wilson et al.

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:46

beliefs should reflect both the knowledge parents have about child-rearing behaviors65 and their confidence in carrying out those behaviors.

This conceptualization of PSE has two important implications. First, PSE is ajudgment based on multiple factors. To feel efficacious, parents must have: (1)knowledge about appropriate parenting practices (e.g., how to soothe a crying infant orset limits for a toddler), (2) confidence in their abilities to enact these practices, (3)

70 belief that their child’s responses are to some extent contingent on their own actions,and (4) belief that others support their efforts (Coleman & Karraker, 1997). Second,PSE can be measured in at least three ways (Jones & Prinz, 2005). General assessmentof PSE asks the extent to which parents feel competent in the parenting role withoutfocusing on particular tasks. Task-related PSE also assesses PSE globally, but does so by

75 having parents assess themselves across a number of specific parenting tasks (e.g.,setting limits, showing empathy). Narrow-domain PSE focuses on efficacy in a singleparenting domain (e.g., discipline). Given the advantages of focusing on specificdomains while still capturing a global assessment, we adopt the task-related approach.

Two literature reviews have explored associations between PSE and parenting80 competence (Coleman & Karraker, 1997; Jones & Prinz, 2005). Jones and Prinz

defined parenting competence as “parenting behaviors, skills, and strategies that havebeen considered to promote positive and adaptive child development and outcomes”(p. 346). Parenting competence includes sensitivity to a child’s cues, displays ofwarmth, consistent responses to child behaviors, and use of nonpunitive discipline as

85 opposed to intrusiveness or detachment, inconsistent responses, and harsh discipline(Hart, Newell, & Olson, 2003). Both reviews emphasize the importance of PSEin promoting parenting competence. PSE also has been inversely associated withparental depression and stress, and positively associated with parental role satisfaction(Jones & Prinz, 2005). Finally, evaluation research has found that parenting training

90 programs that increase positive parenting practices also increase PSE (e.g., Gross, Fogg, &Tucker, 1995).

Although evidence linking PSE and parenting competence is widespread, gapsremain in our understanding of this association. Many of the aforementioned studiesassessed both PSE and parenting competence via self-reports from parents (e.g., Ardelt

95 & Eccles, 2001; Izzo, Weiss, Shanahan, & Rodriguez-Brown, 2000) leaving it unclearwhether parents’ perceptions of their own competence are shared by others. Inaddition, many large-scale studies focus on parenting adolescents rather than youngchildren (e.g., Bogenschneider, Small, & Tsay, 1997; Shumow & Lomax, 2002).Although functions such as connection, regulation, and autonomy granting are

100 important when parenting young children or adolescents (Hart et al., 2003), thespecific parenting practices associated with competence likely vary as children mature.

When only the literature using observational methods to assess parenting youngchildren is examined, findings are mixed. Some studies report significant associationsbetween PSE and parenting competence (e.g., Bohlin & Hagekull, 1987; Teti &

105 Gelfand, 1991) but others report predominately null findings (e.g., Coleman &Karraker, 2003; Corapci & Wachs, 2002). Jones and Prinz (2005) acknowledge thatall of the studies in their review that failed to substantiate a link between PSE and

Parental Self-efficacy and Sensitivity 3

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:46

parenting competence relied on observations of families with young children toassess parenting competence. To account for these seemingly inconsistent results,

110 Jones and Prinz speculated that “some studies may have suffered from a greaterproportion of participants who reported PSE inaccurately, which in turn may havelimited the validity of PSE measurement and reduced the likelihood of detectingeffects” (p. 349).

Assuming that PSE assesses how much confidence a person has in the parenting115 role, the notion that parents could hold inaccurate assessments of their PSE may seem

counterintuitive or even counter-definitional. Of course, parents who are concernedabout being negatively evaluated for admitting a lack of self-confidence might reporthigher PSE than they actually feel (i.e., social desirability bias). Equally problematic,however, parents who have overly idealized views of their competence might genuinely

120 hold high PSE, although their beliefs likely would not correspond with others’perceptions of their knowledge about and ability to enact parenting behaviors thatpromote children’s social and emotional well-being. As noted above, PSE is based onmultiple factors and parents could in fact hold unrealistic or idealized assessments oftheir knowledge about or ability to enact competent parenting practices.

125 One implication of this reasoning is that the relationship between PSE andparenting competence might vary over the range of PSE. PSE in general may bepositively associated with behaviors (e.g., sensitivity) that are reflective of parentingcompetence, but very high levels of PSE, if inaccurate, may be unrelated or eveninversely related with others’ perceptions of parenting competence. If the relationship

130 between PSE and competence is not strictly linear, this could explain why studiesexploring self-reported PSE and observational measures of parenting competencehave yielded inconsistent findings (Jones & Prinz, 2005). Parents of younger childrentend to be younger themselves and newer to the parenting role than parents ofadolescents, and might be especially prone to holding idealized views of their PSE or

135 feeling unsure about reporting insecurities for fear of negative evaluation. This mayexplain why null findings have been most likely to occur in studies of parents withinfants or toddlers.

Research Questions about Parental Sensitivity and PSE

Attachment theory describes the emotional bond—and how this bond forms—140 between a caretaker and child (Trees, 2006). Bowlby (1969/1982) posited that all

young children have an innate and biological desire to remain physically close totheir caregivers which, in turn, keeps them safe. These early bonds have a largedevelopmental impact throughout the child’s life. Through extensive fieldwork andlaboratory observations in multiple cultures, Ainsworth (1967; Ainsworth, Blehar,

145 Waters, & Wall, 1978) showed that parents vary in the degree to which they aresensitive to the signals of their infants and that this sensitivity plays an important rolein the development of secure parent–infant attachment. Attachment security in thiscontext refers to young children’s confidence in their primary caretaker as a secure

4 S.R. Wilson et al.

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:46

base from which to explore the world and a safe haven to which to return when150 distressed (Munz & Wilson, in press).

Parental sensitivity refers to the degree to which parents are “child centered,”meaning that they are aware of their child’s social gestures, expressions, and signalsand respond quickly and appropriately to their child’s signals (especially signs ofdistress; Cox & Crnic, 2003). Insecurely attached infants are more likely to have

155 caretakers who are intrusive or detached. Intrusive caretakers are parent—rather thanchild—centered, in that they impose their agenda on the child despite signals that adifferent activity or pacing is needed. Detached caretakers are emotionallyuninvolved, lack awareness of their child’s signals, and fail to support their child’sexploration (Cox & Crnic, 2003). Consistent with this thinking, a meta-analysis of 21

160 studies (31 effect sizes, N = 1097) found that maternal sensitivity is associated, albeitmodestly, with attachment security, mean weighted r = .24 (De Wolff & vanIJzendoorn, 1997).

Based on this review, we propose a hypothesis and two research questions (RQs)regarding the relationship between parents’ self-reported PSE and observers’ ratings

165 of sensitivity as parents engage in free play with their young children. Given that asizeable literature supports PSE as an antecedent of parenting competence, we predictthat PSE will be positively associated with observer ratings of the level of sensitivity thatparents display during play with their infant/toddler (H1) AQ2. Because overly idealizedPSE may be the equivalent of “too much of a good thing,” we also ask whether there is

170 evidence for a curvilinear relationship between PSE and observer ratings of parents’sensitivity (RQ1). This curvilinear association could result from a subgroup of parentswho self-report high levels of PSE but display low levels of sensitivity during free play.Finally, given the potential importance of understanding how parents who may holdidealized PSE actually behave, we ask in what ways, if any, do parents who report high

175 PSE but are rated as low in sensitivity by observers play with their child differently thanparents who report high PSE and are rated high in sensitivity (RQ2)? To address RQ2,we conduct an in-depth qualitative assessment of a subset of our sample, somethingnoticeably absent from most prior studies of PSE and parenting competence.

Method180 Participants and Research Site

Thirty-eight parents (33 mothers, 5 fathers) participated in this study along with oneof their children. Participants were enrolled in an intergenerational learning programmodeled on Early Head Start in their local community. The program, which targetedparents with limited formal education and family income, was designed to support

185 parents in establishing/pursuing educational goals for themselves and their children.Children could participate in the program from infancy up to 36 months. Amongseveral components, the program offered a well-validated parenting curriculum (theIncredible Years; Webster-Stratton & Hammond, 1997). All data reported weregathered during each family’s first week in the program.

Parental Self-efficacy and Sensitivity 5

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:46

190 The mean age of parents was 23.31 years (SD = 4.80, range: 16–40). About one-third of the parents (31%) were between 16 and 20 years and only two were 30 yearsor above. The average age of the child was 14.26 months (range: 2–36). Forty-fivepercent of the children were infants (2–12 months), 45% were young toddlers (13–24months), and 10% were older toddlers (25–36 months). In terms of ethnicity, 18% of

195 parents selected the option of “Hispanic/Latino/a” and 82% selected “non-Hispanic/Latino/a.” In terms of race, 42% of parents selected “White,” 24% “Black,” 3%“Asian,” 10% “Biracial or Multiracial,” and 21% “Other.” The sample includedparents with varying degrees of education: 34% were currently enrolled in 10th, 11th,or 12th grades, 29% had a high school diploma or GED AQ3, 29% had completed some

200 college or an associate degree, 3% had a four-year degree or higher, and 5% did notrespond. Most parents spoke English as their primary language (92%) followed bySpanish (5%) and Chinese (3%). Regarding marital status, 79% were single parents,16% were not single parents, and 5% did not answer the question. The sampleincluded families with one child (40%), two children (18%), and three or more

205 children (32%) at home (10% did not respond to the question).All parents in the program were pursuing an education goal. Aside from attending

school, 53% of participating parents had no paid work, 27% worked part time, 7%worked full time, and 13% did not answer. The average annual family income was$11,046.42.

210 Procedure

Parents were recruited for a longitudinal evaluation study when they enrolled in thelearning program. Following procedures approved by the local university IRB AQ4, one ofthe researchers attended the enrollment meeting and briefly explained the study’spurpose and procedures. The researcher stressed that participation in the evaluation

215 study was voluntary, and that families could enter the program regardless of whetherthey participated in the evaluation research. The vast majority (more than 90%) offamilies enrolled in the program consented to participate in the evaluation study.Those who agreed were scheduled for a Time 1 data collection session (whichincluded the measures reported here) within the next week.

220 Data collection sessions, which occurred in the same building where infants/toddlers attended early childhood education classes, lasted 60–90 minutes. Duringthis session, parents completed questionnaires and were videotaped while engaged infree play with their child. If a parent had two children in the program, the parentparticipated along with their oldest child.

225 Parents completed five questionnaires about their experiences of being a parentand their perceptions of their child, one of which was a measure of their PSE. Parentsand children also were videotaped playing together for 15 minutes with a basket ofage-appropriate toys (e.g., books, blocks, stuffed animals, and music toys). Theprecise contents of the basket varied depending on the age of the child but each

230 contained similar objects. For example, each basket contained a book: a soft plasticpicture book for infants, a board book for young toddlers, and a book promoting

6 S.R. Wilson et al.

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:46

reading readiness for older toddlers. The basket was placed on a mat in the room andparents were told that they and their child could play with one or more of the toys inany order they chose. A video camera was set on a tripod at the edge of the room. A

235 cord ran from the camera to a monitor where the researcher sat (behind a bookcase,out of view of the parent and child) to ensure that the equipment functionedproperly. Parents were asked to keep their child on the mat as much as possibleduring the play period so that they would be within the range of the camera. Theorder in which parents completed the questionnaires and were videotaping varied

240 across families depending on when the data collection occurred (e.g., drop-off versuspick-up).

In most cases, only mothers attended the meeting in which families were recruitedfor data collection. When a father or father figure was involved in caring for thechild, he also was invited to participate in the study. In those cases where two parents

245 did participate, they both completed questionnaires and then decided which one ofthem would be videotaped playing with their child. Only data from that parent wereanalyzed in this study.1 Parents received $20 for completing both parts of the datacollection and $15 for completing only the questionnaires.

Measurement

250 Parental self-efficacy (PSE). Parents responded to the 53-item Self-Efficacy forParenting Tasks Index—Toddler Scale (SEPTI-TS; Coleman & Karraker, 2003). TheSEPTI assesses parents’ confidence in their knowledge and ability to perform a rangeof specific tasks involved in parenting young children. Specifically, the SEPTI tapsperceived efficacy within seven sub-dimensions: (1) emotional availability (e.g.,

255 “When my child needs me, I am able to easily put aside whatever else I may bedoing”), (2) nurturance/empathetic responsiveness (e.g., “My toddler knows that Iunderstand when his/her feelings are hurt”), (3) protection (e.g., “I always make sureI can see my child in order to make sure he/she does not get hurt”), (4) discipline/limit setting (e.g., “I have trouble getting my child to listen to me” [reverse scored]),

260 (5) play (e.g., “I am able to get actively involved when playing with my child”), (6)teaching (e.g., “Helping my child learn colors, names of objects, etc. is not one of mystrongest points” [reverse scored]), and (7) instrumental care/routines (e.g., “I havebeen able to establish a daily routine with my infant/toddler that feels comfortablefor both of us”). Parents responded to items using a 6-point scale ranging from

265 1 = “strongly disagree” to 6 = “strongly agree.” Higher scores indicate higher self-efficacy. Several SEPTI items were edited slightly to improve the readability of thescale for parents with lower literacy levels. For example, the item “Providing a safe,hazard-free environment for my child is very difficult for me” was revised to“Providing a safe home, where accidents won’t happen, is very difficult for me.”

270 After deleting seven items with low item-total correlations, internal consistency forthe remaining 46-item scale was excellent (α = .93). As in Coleman and Karraker’s(2003) original report, reliabilities for some subscales were modest, with alphas rangingfrom .65 to .82. All analyses reported here use the total SEPTI scale (i.e., 46 items). In

Parental Self-efficacy and Sensitivity 7

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:46

support of the scale’s concurrent validity, Coleman and Karraker found that mothers’275 scores on the SEPTI were correlated (r = .48) with the 7-item “efficacy” subscale of the

Parenting Sense of Competence Scale (Johnston & Mash, 1989), a global or domain-general measure of PSE.

Observer ratings of parental sensitivity. The Qualitative Ratings for Parent–ChildInteraction (QRPCI) scales (Cox & Crnic, 2003) were used to rate the degree to

280 which parents displayed sensitivity during the 15-minute play period with their child.Because low levels of sensitivity may reflect that parents are either highly intrusive ordetached, observers rated these two dimensions as well. QRPCI ratings have beenassociated with other theoretically related variables such as children’s attachmentsecurity and maternal depression (e.g., Barnett, 2008; Frosch, Cox, & Goldman, 2001;

285 Semple, Mash, Ninowski, & Benzies, 2011).Coders, masked to parents’ self-efficacy scores, rated each of the three dimensions

of parental behavior. Sensitivity was judged based on whether parents acknowledgedtheir child’s affect and vocalizations, displayed good timing paced to their child’sinterest, picked up on their child’s interest in toys or games, and responded quickly

290 and appropriately to signs of child distress (e.g., cries). Intrusiveness was rated basedon whether parents offered a continuous barrage of stimulation, took objects awayfrom the child, manipulated the child’s body in an intrusive manner, or failed toquickly alter their own behavior when their child turned away. Detachment wasjudged based on whether parents rarely talked to or made eye contact with their

295 child, presented toys without showing how to use them, failed to respond to theirchild’s vocalizations, or behaved in a mechanical fashion. The QRPCI scales focus ona parent’s verbal and nonverbal behaviors, including how the parent responds to thechild’s signals (e.g., what the parent does when a toddler tries to get a toy that is outof reach).

300 Each 15-minute play period was broken into five 3-minute segments. Coders ratedthe degree of sensitivity, intrusiveness, and detachment a parent displayed duringeach segment. All three dimensions were rated on a 1–5 scale where 1 = not at allcharacteristic of the parent, 2 = minimally characteristic, 3 = somewhat characteristic,4 = moderately characteristic (e.g., occurred more often than not during the 3-minute

305 segment), and 5 = highly characteristic. For each dimension, the QRPCI includesexamples of what parenting behaviors should be scored at each of the five levels forparents interacting with either infants or toddlers.

To assess inter-coder agreement, intraclass correlations were computed after threecoders had rated play periods for 20 parents (N = 99, 3-minute segments; one

310 videotape lacked sound for the first three minutes). Reliabilities were good asintraclass correlations were .78 for sensitivity and .77 for both intrusiveness anddetachment. Reliabilities were again calculated for two additional coding phases andacceptable inter-coder agreement was maintained.2 The three coders’ ratings for each3-minute segment were averaged and then these combined scores, in turn, were

315 averaged across the five 3-minute segments composing each play period to get anoverall 1–5 point rating of that parent’s sensitivity, intrusiveness, and detachment.

8 S.R. Wilson et al.

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:46

Qualitative Follow-up: Participants and Analysis

To examine how parents who reported high PSE but were rated low in sensitivitymay have played with their child differently than those who reported high PSE and

320 were rated high in sensitivity, we conducted a qualitative analysis with two groups ofparents (n = 4 each) from the larger sample who fit these profiles. Because all eightparents were mothers, we use the shorthand “LSM” (low-sensitivity mothers) and“HSM” (high-sensitivity mothers) to indicate these two groups. To be selected, aparent had to: (1) fall at or above the 75th percentile in terms of PSE and (2) either

325 below the 25th percentile (LSM) or at or above the 75th percentile (HSM) terms ofsensitivity. Within these constraints, we selected families who were as demograph-ically similar as possible. By choosing two groups who both had very high PSE butfell at opposite ends of the scale in terms of sensitivity, we hoped to identityqualitative distinctions that were useful in interpreting quantitative findings and

330 depicting what idealized levels of PSE look like.Each author initially was responsible for analyzing two interactions, one in the LSM

category and the other one in the HSM category. Researchers were alert to contrasts andsimilarities between videotaped play periods, with an emphasis on understandingspecific behaviors that may have led observers to rate LSMs as insensitive (e.g., behaviors

335 that communicate lack of awareness or responsiveness to a child’s signals). Analysisproceeded with authors transcribing and bringing their assessments of both unique andsimilar behaviors to the group, and these assessments were then scrutinized by the wholegroup through repeated viewings and the detailing and correction of transcripts.

Ultimately, the types of verbal and nonverbal actions that occurred within the340 category of HSM or LSM were assembled under descriptive headings (e.g., “don’t”

imperatives, restricting access to toys, and maintaining mutual gaze with child). Oncethis set of actions was identified, researchers returned to the task of looking forbehaviors across categories. HSM tapes were inspected for instances of LSM behaviorand vice versa. We searched for actions that were: (1) performed by LSMs but not

345 HSMs, or (2) performed by both groups but in noticeably different ways. For thefinal analysis, two of the authors retranscribed all instances presented here, adaptingthe system of transcription typically employed in conversation analysis (Sacks,Schegloff, & Jefferson, 1974). Although the authors were not masked to whichmothers were LSM versus HSM, the detailed nature of this transcription system does

350 constrain the researcher from importing a-priori assumptions into the analysis.

Results

Preliminary Analyses

Descriptive statistics for PSE and ratings of parental sensitivity, intrusiveness, anddetachment during the play period for the sample as a whole appear in Table 1.

355 Parents’ average total score on the SEPTI was 237 points; dividing this summed scoreby the total number of items (46) reveals that parents’ average score was 5.15 on the1–6 scale. This mean score is similar to but slightly higher than Coleman and

Parental Self-efficacy and Sensitivity 9

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:46

Karraker (2003) who report (M = 5.04) for their original sample of N = 68 mothers.Parents in the current study, on the whole, self-reported high levels of PSE at the

360 time they entered the intergenerational learning program.Parents as a group varied substantially in terms of sensitivity, with ratings across

the 15-minute play period ranging from under 2.0 to nearly 5.0 (see Table 1). Mostparents displayed low to moderate levels of intrusiveness, though some were abovethe scale midpoint. Parents displayed very little detachment, perhaps reflecting that

365 the play period was short and parents were being videotaped. As would be expected,observer ratings of a parent’s sensitivity were inversely associated with ratings of theparent’s intrusiveness, r(36) = −.77, p < .001, as well as detachment, r(36) = −.26,p = .12, although the latter correlation was not statistically significant.

Descriptive statistics for two groups of mothers (n = 4 LSM, n = 4 HSM) selected for370 the qualitative follow-up analysis are shown in Table 2. For each group, three of the four

mothers fell above the 75th percentile in terms of the overall sample’s level of PSE,whereas the fourth fell just below the 75th percentile. The two groups differeddramatically in terms of their rated sensitivity. The four LSMs fall between the 5th and18th percentile in terms of the overall sample’s level of sensitivity, whereas all four

375 HSMs fall at the 75th percentile. The two groups are demographically similar in termsof maternal age and race as well as child age and sex.

Correlations between parents’ self-reported total PSE, observer ratings of parentalsensitivity, and several demographic variables (child age and sex, parent age andeducation) for the sample as a whole appear in Table 3. Other demographic factors were

380 not analyzed due to uneven distributions (e.g., for parent sex, participants include 33mothers but only 5 fathers). None of the four demographic variables were related toobserver ratings of parental sensitivity. Child sex was marginally associated withparents’ self-reported total PSE, reflecting that parents of girls (M = 243.65 SD = 18.53,n = 18) reported slightly higher self-efficacy than parents of boys (M = 230.84,

385 SD = 26.43, n = 20). Child age, parent age, and parent education were not associatedwith PSE. Given these findings, analyses reported below control only for child sex.

PSE and Observer Ratings of Parental Sensitivity

Hypothesis 1 predicted a positive relationship between PSE and parental sensitivity.The relationship between PSE and sensitivity was not significant (see Table 3). As a

Table 1. Descriptive statistics for PSE and sensitivity for the overall sample.

Measure Min Max M SD Skewness Kurtosis

PSE (SEPTI total) 164 276 236.91 23.63 −0.51 1.20Sensitivity rating (avg.) 1.93 4.73 3.54 0.63 −0.56 0.05Intrusiveness rating (avg.) 1.00 3.47 2.05 0.68 0.64 −0.70Detachment rating (avg.) 1.00 2.40 1.36 0.38 1.42 1.61

N = 38 parents. Scores on the SEPTI measure of PSE can range from 46 to 276; sensitivity, intrusiveness, anddetachment were rated on a –5 scale.

10 S.R. Wilson et al.

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:46

390 linear relationship, the degree to which parents self-reported feeling knowledgeableabout and capable of accomplishing various parenting tasks was not associated withobserver ratings of their sensitivity while playing with their infant/toddler. Thisfinding runs contrary to Hypothesis 1.

To address RQ 1, a scatterplot was examined to investigate the possibility of a395 nonlinear relationship between self-reported PSE and ratings of parental sensitivity.

As can be seen in Figure 1, the negative association between PSE and ratings forparental sensitivity occurs primarily as one moves from moderate to high PSE scoresin terms of the levels of self-efficacy reported by parents in our sample. Whenmoving from low to moderate levels of efficacy, the association between PSE and

400 sensitivity is slight but positive (as predicted in H1).To provide a more rigorous test of curvilinearity, a hierarchical multiple regression

analysis was conducted. Parental sensitivity served as the criterion variable. Child sex

Table 2. Descriptive statistics for eight mothers in qualitative follow-up analysis.

Mom ID

Momage

(years)

Momrace/

ethnicity

Childage

(months)Childsex

MomPSE

(SEPTItotal)

Momsensitivity(rated)

Momintrusive(rated)

Momdetach(rated)

LSM4 24 Latina 20 Female 248 2.93 2.8 1LSM8 16 Black 7 Female 274 2.27 3.4 1LSM11 22 White 7 Male 253 2.73 1.87 2.4LSM33 25 Black 18 Female 276 2.4 2.8 1.87M LSM 21.75 1W1L2B 13.00 3F1M 262.75 2.58 2.72 1.57HSM6 20 White 6 Male 259 4 1.67 1.27HSM7 25 Latina 18 Female 241 4 1.23 1.53HSM17 18 Black 4 Male 260 4 1.8 1.13HSM32 25 Black 19 Female 269 4 1.45 1.6M HSM 22.00 1W1L2B 11.75 2F2M 257.25 4.00 1.54 1.38

Total scores on the SEPTI measure of PSE could range from 46 to 276 points. Sensitivity, intrusiveness, anddetachment were rated on –5 scales (averaged across coders and 3-minute segments).LSM, low-sensitivity mom;HSM, high-sensitivity mom.

Table 3. Correlations between parenting sensitivity, PSE, and demographics.

Parental sensitivity(ratings)

PSE (SEPTItotal)

Childsex

Childage

Parentage

Parenteducation

PSE −.23 1.0C sex .08 .27† 1.0C age −.06 −.23 −.03 1.0P age .06 −.16 .01 .45* 1.0P education .19 −.15 −.19 .43* 26 1.0

N = 38 parents. *p < .0, †p < .10. Parent education was measured using five ordinal categories varying from 1 =9th grade or less to 5 = four-year college degree or higher. For child sex, 1 = male, 2 = female.

Parental Self-efficacy and Sensitivity 11

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:46

(1 = male, 2 = female) was entered as a control variable at Step 1. Scores for the lineareffect of parent self-reported total PSE were entered at Step 2, and then scores for the

405 quadratic effect of PSE were entered at Step 3 (see Table 4). To reduce multi-collinearity, PSE raw scores were centered for the linear effect, and then centered PSEscores were squared to create the quadratic effect. Consistent with the correlationalanalyses, the linear effect of PSE on sensitivity was not statistically significant whenentered at Step 2, although it became significant after partialling out variance

410 explained by the quadratic effect at Step 3. More importantly, the quadratic effect wasstatistically significant at Step 3 even after controlling for the linear effect and childsex.3 This regression analysis indicates that the association between PSE andsensitivity is curvilinear rather than strictly linear. The quadratic effect explained10% of the variance in parental sensitivity.

415 Qualitative Follow-up Analyses

RQ2 asked in what ways parents who reported high PSE but were rated low insensitivity by observers played with their child differently than parents who reportedhigh PSE and were rated high in sensitivity. We identified four behaviors typical ofLSMs but not HSMs. First, LSMs attempted to introduce new toys when their child

420 already was engaged with a different toy. Examples included LSMs putting a new toy intheir child’s line of vision, placing a different toy physically in front of their child, or

Figure 1 Scatterplot for scores on PSE (SEPTI total scores) and parental sensitivity.

12 S.R. Wilson et al.

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:47

shaking a new toy when their child already was looking at, touching, or manipulating adifferent toy. Coders may have perceived such actions as lower in sensitivity becausethe mother either was not aware of her child’s signals or insisted on her own activity

425 choices rather than building on her child’s current activity choice.Excerpt 1 involves LSM 11 and her infant son (7 months) playing on a mat after

the mom has taken several toys (ring stacker, foam blocks) out of a plastic basket. Atthe start, the infant is holding onto the end of the basket with two hands—pulling itup to him and putting the corner into his mouth. He then releases it from his mouth

430 but is still holding it in his hands and moving it back and forth when his mom triesto introduce a different toy.

Excerpt 1 (LSM 11)01 M: Look Jona↑h (.) look 435As she says “Jonah,” M picks up a ring

stacking toy from behind the basket andmoves it into C’s line of vision. M shakesthe toy as she says “look” thesecond time.

440 02 (8.0) M takes the top ring off the stack andshakes it (it rattles loudly). C briefly turnshis gaze toward the ring. Before M cantake off a second ring, the infant turns

445back to the basket.03M: tsk (.) You got all the↑se toys hereand you wanna chew on that basket

As M puts top two rings back on stack,she makes a “tsk” sound and then

450delivers turn 304 (4.7) M starts to move other toys out from

behind basket and puts them betweenthem on the mat

455

Table 4. Hierarchical regression of parental sensitivity onto child sex, PSE (Total SEPTIscores), and squared total SEPTI scores.

Predictor β Total R2 R2 change

Step 1Child sex .08 .01Step 2Child sex .15Total SEPTI scores (Linear) −.27 .08 .07Step 3Child sex .12Total SEPTI scores (Linear) −.36*Total squared SEPTI scores (Quadratic) −.33* .18† .10*

N = 38 parents. For child sex, = male, 2 = female. *p < .05, †p < .10.

Parental Self-efficacy and Sensitivity 13

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:47

05M: What’s up with that? (.) Look yougotta boo:k you got blo:cks, you gottapho:ne, you got one of these thingies

460 whatever they are (.) and you’re moreinterested in a ba↓sket

When M says “boo:k” she holds up thebook, but C does not turn his gaze awayfrom the basket. As M points to other

465toys, C does reach out and touches afoam block with his hand. M does notnotice because she is looking down ateach toy she lists rather than at C.

470 For reasons we cannot know from the data, LSM 11 does not appear to view thebasket as an appropriate toy and hence attempts to redirect her child’s interest fromit to other possible activities. When she is unable to engage her son in a differentactivity, LSM 11 sounds exasperated as she lists other toys with which he might play(turn 05). She also fails to notice one moment when her son does direct his attention

475 to a toy that she lists (turn 05).HSMs occasionally attempted to introduce new toys when their child was

currently engaged with a different toy, but they did so indirectly and withoutforcing their activity choice. HSM 32 tried to introduce a ring stacking toy when herchild was still interested in a foam block. When her child pushed the ring toy away,

480 HSM 32 said “oh, you still want to play with the block?” She then played with thering toy herself while watching her daughter play with the foam block, at one pointasking if her daughter needed help. After her daughter’s gaze moved to the ringstack toy, HSM 32 put it beside her daughter without moving the block. Sheremoved one ring and then praised her child when the daughter began to remove

485 rings from the stack.Second, LSMs physically blocked their child’s access to toys even when their child

was demonstrating interest by reaching for or gazing at the toy. Sometimes thisoccurred in the form of an extended “keep-away” game that their child showed noevidence of enjoying, whereas in other cases LSMs would move toys their child

490 currently was playing with out of the child’s reach while trying to introduce a newtoy. Coders may have perceived these behaviors as insensitive because they promotethe mother’s rather than the child’s agenda by forcing the mom’s activity choices. Insome cases the mom’s “pacing” was not in sync with her child’s signals.

In Excerpt 2, LSM 8 places her infant daughter (7 months) in a sitting position on495 the mat, and kneels down facing her. LSM 8 shakes a rattle in front of her daughter

to get the infant’s attention, at which point the infant looks up and reaches for therattle with both hands. LSM 8 pulls the rattle back away from the infant, then movesthe rattle to the infant’s left and right, then up above the infant, then back to infant’sleft and right, shaking it the entire time. At this point, LSM 8 puts the rattle back

500 within the infant’s reach, but when the infant moves to grasp it her mom pulls itaway. LSM 8 repeats the process a second time. At that point LSM 8 sits the rattle onthe mat between the infant’s legs, then turns her head and reaches into the basket toremove a stuffed tiger. Transcription starts at the point where the mom puts downthe rattle.

14 S.R. Wilson et al.

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:47

Excerpt 2 (LSM 8)01 M: You is very smart Kiara. M is looking away from C towards a basket

with toys when saying this. M reaches into the510basket and pulls out a stuffed tiger. C is

holding the rattle with both hands.02 M: Look Kiara. (.) Kiara look.(0.8) Look (0.5) ↑Stop↑. Look.

As M says “Kiara look” a 2nd time, M puts her515left hand on C’s right shoulder and pushes C

back away from the rattle. At “stop” M usesthe stuffed tiger to move the rattle away fromC with her right hand while continuing topush C back with her left hand. At the final

520“look” M puts the stuffed tiger into the C’shands, at which point C leans back andholds it.

In this excerpt, the infant tracks her mother’s movement of the rattle and reaches525 for it multiple times. Although LSM 8 may be withdrawing the rattle as a “keep-

away” game, the child does not laugh or give any other indication of enjoying the“game.” After LSM 8 finally allows her daughter to grasp the rattle, she immediatelyattempts to introduce a new toy rather than letting her daughter play with the rattle(turn 02). LSM 8 physically pushes the rattle out of the way while putting the new toy

530 tiger into her child’s arms.HSMs occasionally restricted access to toys as well, but they did so for shorter

periods of time and were more attentive to their child’s signals. For example, HSM 6briefly played “keep-away” with her 6-month-old son by withdrawing a rattle after hereached for it, but she quickly put the rattle back in his reach, allowed him to touch

535 it, and then rotated it to facilitate the infant being able to grasp the handle of therattle while commenting on his interest in it.

Third, LSMs physically manipulated their child, which involved holding and/ormoving the child in a way such that the child did not have control of his/her bodyand did not show evidence of enjoying (e.g., no smiling or laughter). Often, these

540 examples included what the literature has described as parental physical negativetouch, defined as “any physical touch that is intended to be antagonistic, aversive,hurtful, or restrictive of the … child’s activity” (Borrego, Timmer, Urquiza, & Follett,2004, p. 899). Examples of this occurred when LSMs made their child dance like a“puppet,” held their child in place while scolding them (e.g., for putting objects in

545 their mouths), or roughly moved their child to a different position (e.g., from layingto sitting). In other cases, physical manipulation involved LSMs holding or touchingparts of the child’s body such as tickling a child for an extended period of timewithout the child smiling/laughing or grasping the child by the wrist to force thechild to explore an object in which the child was not showing interest.

550 Excerpt 3 begins with LSM 4 sitting on a mat holding a music toy that is playing asong. Her daughter (20 months) is sitting on the mat, perpendicular to her motherholding a stuffed Mickey Mouse toy in both hands. The child moves the toy along

Parental Self-efficacy and Sensitivity 15

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:47

with the music as if it were dancing. LSM 4 suggests to her daughter that she shouldstand up and dance with Mickey.

Excerpt 3 (LSM 4) AQ5

01 M: Get up As she says “Get up” M reaches under C’s arms withboth hands and tries to lift C into a standing position

560(C is facing away from M).02 (1.0) M keeps her hands under C’s arms, trying to lift her.

C’s body goes limp.03 M: ↑Get up↑ 565As she says “get up” again, C bends her knees, trying

to sit down in M’s lap. M keeps her hands under theC’s arms, still trying to lift C to a standing position.

04 (2.2) Holding C under the arms, M moves C’s body back570and forth, as if dancing. C’s body remains limp as M

shifts C from left to right.05 M: ↑Get up↑ = ↑Get up↑(.) ↑Come on↑

As M says “Get up” for the second time this turn, C575sits down in the mother’s lap and the Mickey doll

drops out of her hands.06 (1.0) M reaches out for Mickey07 M: Here you go 580M picks up Mickey from the mat and gives it to C as

she says “Here you go.”08 (1.2) Sitting on her mom’s lap, C begins to move Mickey up

and down, as if dancing585 09 M: Up As M says “up,” she again grasps C under the arms

and tries to lift C to a standing position10 (1.0) C bends her knees, letting her body go limp.

590 11 M: No? M says “no?” as C remains limp.

In this excerpt, LSM 4 continues to physically manipulate despite many signs thatthe daughter is not participating in or enjoying the “dancing,” such as the daughter

595 going limp in turns 02 and 10, as well as sitting back down on her mother’s lap inturn 05. The mother’s “No?” in turn 11 indicates that she recognizes her daughter’slack of participation, but this recognition is verbalized only after she ignores multiplecues of disinterest from the daughter.

None of the four HSMs engaged in physical negative touch. HSMs would600 occasionally hold their child under the arms, but this was done to facilitate an

activity the child initiated (e.g., HSM 32 holding her daughter so the toddler couldtouch her mother’s face without falling over).

Fourth and finally, LSMs violated their child’s proximal space when they touchedtheir child with objects or put objects in their child’s face even after their child

605 displayed signs of discomfort such as turning or moving away. In these examples,LSMs invaded their child’s personal space with objects rather than physicallytouching or holding their child. Coders may have perceived such behaviors asinsensitive because LSMs either failed to show awareness of or ignored their child’ssignals that the activity was not enjoyable.

16 S.R. Wilson et al.

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:47

610 Excerpt 4 starts as LSM 33’s 18-month-old daughter is rolling a “popper” toy backand forth on the mat while her mom holds her by the back of her shirt. The toddleris smiling while rolling the popper. After several rolls, the child drops the popper andfalls to her knees, bumping the ring stack toy. At this point, the daughter picks up thering stacker. LSM 33 picks up the popper and begins to run it into the child’s leg

615 while making collision sounds (turn 03).

Excerpt 4 (LSM 33)01 M: Wo:: As M says “wo::” she reaches out with her hand to

620prevent C from falling forward.02 (7.0) M reaches for the popper toy that C had dropped. As

C is looking at a ring stack toy, M bumps C in thethigh softly 5 times with the popper. After the 5th

625bump, C turns her head to the left and looks at thepopper briefly (1 sec).

03 M: pwoo:: (.) >pwoo:pwoo: pwoo: pwoo:<

M says the first “pwoo” as C turns her gaze towards630the popper. By the second “pwoo” C turns her gaze

back to the right, away from the popper. By the 4th“pwoo” C puts her hands on the mat. C is no longersmiling.

04 (1.8) 635C pushes herself up to a standing position.05 M: pwoo:::: M pushes the popper into C’s left hand as C is nearly

standing.06 (4.0) 640C turns away from her mom and walks off the mat.

M bumps C with the popper in the back of C’s foot asC is turning and moving away.

07 M: Come here Said as the child walks off the edge of the mat.645

Especially evident is how LSM 33 continues bumping her daughter after thetoddler had stopped smiling (turn 03), turned her body away (turn 06), and walkedoff the mat (turns 06–07).

HSMs occasionally invaded their child’s personal space with toys but they650 monitored the child’s reaction and stopped if their child displayed signs of not

enjoying the activity. For example, HSM 17 put a Mickey Mouse toy in her infant’sson’s face while making a “raspberry” noise (“pblt::”) and then removed it. The infantsmiled while looking back and forth between his mother and the toy. HSM 17 movedthe toy toward her son’s face a second time, but when the infant turned his gaze to

655 the right she immediately withdrew the toy and did not repeat the action.

Discussion

This study examined the relationship between self-reported PSE and observer ratingsof sensitivity as low-income parents who were entering an intergenerational learningprogram completed an unstructured play period with their infant or toddler.

660 Although PSE was expected to be positively associated with sensitivity, a curvilinear

Parental Self-efficacy and Sensitivity 17

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:47

association between self-reported PSE and observers’ ratings of sensitivity actuallywas discovered in this sample. The quadratic effect was statistically significant andsubstantial even after controlling for any linear association as well as for child sex.Although the association between PSE and sensitivity is curvilinear, the suppressed

665 linear effect indicates that the quadratic parabola is slightly inclined (see Step 3 inTable 4). Put simply, the positive association between PSE and sensitivity runningfrom low to moderate levels of self-efficacy in our sample is weaker than the negativeassociation running from moderate to high levels of self-efficacy.

One contribution of this study is that we formally tested for and found a curvilinear670 association between PSE beliefs and parental sensitivity. Had we only assessed the

linear association, our null findings would have been consistent with some priorresearch; indeed, Jones and Prinz (2005) note that null results typically have occurredin samples with young children that employ observational measures. Prior studies havetended to attribute null findings to the limits of assessing parenting competence based

675 on only short periods of observation in artificial settings (e.g., Coleman & Karraker,2003, pp. 140–141); however, we were able to detect a significant quadratic effect in oursample even with a short observation period.

The curvilinear association appears to reflect that a subset of parents from oursample self-reported very high PSE yet were rated insensitive by observers when they

680 engaged in free play with their child. This curvilinear pattern does not imply thatevery parent in our study who self-reported high PSE was rated as low in sensitivity.What the pattern does reflect, however, is that the majority of parents in our samplewith moderate PSE were rated high in sensitivity, whereas a group at the upper endof PSE were rated as very low in sensitivity (see Figure 1).

685 A second contribution is our use of qualitative methods to gain greater insightabout mothers who reported high PSE but who were rated as low in sensitivity byobservers. We identified four behaviors that typify LSMs; HSMs either did notperform these behaviors or enacted them in qualitatively different ways. Twobehaviors (introducing new toys when the child already was occupied, restricting the

690 child’s access to toys) involved how the mother and child negotiated which toys toplay with at a specific point in time, whereas the other two (physically manipulatingthe child, proximal space violations) involved mothers persisting in actions thatinvaded their child’s space despite signals the child was not participating in orenjoying the activity. Through these behaviors, LSMs displayed lack of awareness or

695 inappropriate responses to their child’s signals by insisting on their own activitychoices rather than following their child’s lead. LSMs also seemed out of sync whenthey introduced new toys too quickly or terminated activities their child did notenjoy too slowly. Our qualitative analyses offer greater insight into what insensitivityactually looks like when enacted by mothers with high self-efficacy in the context of

700 free play. By analyzing detailed transcripts, we move beyond prior studies that havecoded for sensitivity using the QRPCI but offered only impressionist summaries ofdifferences between more and less sensitive parental responses during play (e.g.,Semple et al., 2011).

18 S.R. Wilson et al.

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:47

After seeing the qualitative findings, it is difficult to argue that observer ratings of705 low sensitivity for LSMs are simply a function of cultural differences in what

constitutes sensitive or competent parenting. LSMs and HSMs are nearly identical interms of demographic profiles (see Table 2). In addition, the conceptual frameworkunderlying the QRPCI coding system—Ainsworth and colleagues’ (1978) work onmaternal sensitivity and its contribution to infant attachment—itself was based on

710 fieldwork with culturally diverse samples. In sum, it is clear that at least in thissample, a subset of parents reported high PSE but displayed behaviors that priorliterature would characterize as indicative of parental incompetence in the context offree play. This subset of mothers was rated as among the least sensitive of all parentsin our sample, which helps account for the curvilinear relationship between PSE and

715 sensitivity (see Figure 1).This finding has both theoretical and practical implications. Theoretically, the

findings challenge any simple association between PSE and parenting competence,and suggest the need for future research clarifying the nature of idealized orunwarranted PSE. Bandura (1986) recognized that self-efficacy judgments might not

720 be associated positively with behavioral performance under some conditions, such aswhen individuals make faulty self-assessments due to insufficient experience in aparticular domain or personal factors that distort efficacy perceptions. Suggestive is astudy by Conrad, Gross, Fogg, and Ruchala (1992), who found that maternalconfidence and mother’s knowledge of child development interacted in their effect on

725 the rated quality of mother–child interactions during a structured teaching task. Formothers scoring in the upper third in terms of child development knowledge,maternal confidence was positively associated with the quality of mother–toddlerinteraction, whereas for mothers scoring in the lower third, maternal confidenceactually was associated inversely with interaction quality.

730 One future direction, then, is to distinguish various elements underlying PSE suchas with distinct assessments of parental knowledge of appropriate childcareresponses, confidence in their parenting abilities, belief that they and not just theirchild drive patterns of interaction, and judgment that others support or interfere withtheir parenting efforts. A second direction would be to explore factors, both

735 demographic (e.g., young parental age) and cognitive (e.g., perceived causes ofcaregiving success or failure; Bugental & Happaney, 2000), that may characterizeparents with high but idealized PSE. Given that self-efficacy is an importantcomponent of several communication theories (e.g., the Theory of MotivatedInformation Management, Afifi, 2010; the Extended Parallel Process Model,

740 Maloney, Lapinski, & Witte, 2011), the notion of idealized or unwarranted self-efficacy may have broader theoretical relevance as well.

With regard to practical implications, parenting class instructors and programdirectors may struggle with how to reach all types of families and work with parentswho believe they will not benefit from participating in a parenting program.

745 Participants in this study were entering an intergenerational learning program wherethey were required to attend parenting classes using the Incredible Years curriculum(Webster-Stratton, & Hammond, 1997). Like the most effective parenting programs

Parental Self-efficacy and Sensitivity 19

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:47

(Kaminski, Valle, Filene, & Boyle, 2008), the Incredible Years focuses on creatingpositive parent–child interactions by teaching parents to follow their child’s interest

750 during free play. At the start, not all parents were equally receptive to the curriculum.The fourth author encountered this challenge while co-teaching the parenting class.For example, one mother with four children repeatedly commented on how shebelieved the class was beneficial for new parents but was less useful for parents likeher who already had multiple children. Despite these sentiments, the same mother

755 expressed concerns about her children’s behavior and sleeping habits. Idealized PSEmay have hindered this mother from seeing how the curriculum might help addresssome of her parenting concerns without undermining her credibility as a parent.Knowledge of the curvilinear association between PSE and sensitivity might helpinstructors to understand possible sources of resistance to learning and develop

760 strategies for reaching parents who have high levels of confidence but lack knowledgeof certain competent parenting practices. Programs certainly do not want toundermine parents’ PSE, since lacking confidence in the parenting role is associatedwith its own set of problems such as depression, parenting stress, and lowerparenting satisfaction (Jones & Prinz, 2005). Our findings also suggest that

765 evaluations of parenting programs would be wise to supplement pre-post measuresof PSE with assessments of parenting behavior and child outcomes.

Several limitations on our findings should be considered. First, the sample size issmall (N = 38), hence, confidence in the curvilinear effect would be buttressed byreplication with a larger sample. Second, the sample included both mothers and

770 fathers, but not enough fathers to examine parent gender differences. Third, childrenranged in age from 2 to 36 months. Although the QRPCI provides guidelines forrating parental sensitivity with both infants and toddlers, the precise markers of (in)sensitivity vary with child age. Fourth, parents were observed interacting with theirinfant or toddler for a short time period while engaged in a single activity (free play).

775 Making judgments about parenting competence from a single observational sessionshould be done cautiously, although brief periods of observation often are sufficientto distinguish differences between parents (e.g., Wilson, Rack, Shi, & Norris, 2008).Fifth, although the qualitative analyses identified concrete behaviors that distin-guished our LSM and HSM subgroups, we did not quantitatively code for these

780 behaviors in the larger sample so group differences should be interpreted withcaution. Finally, parents knew that they were being videotaped while playing andhence were likely displaying behaviors they believed were indicative of goodparenting. This may explain why parents rated as low in sensitivity in this studyusually displayed intrusive rather than detached behaviors.

785 In sum, our findings suggest that the association between PSE and parentingcompetence is more complicated than typically has been recognized. A morenuanced understanding of the connection between PSE and parenting sensitivityalso has important implications for parenting programs such as the one offered bythe intergenerational learning center studied here.

20 S.R. Wilson et al.

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:47

790 Notes[1] Because only one parent was videotaped playing with their child even when two attended the

intake, a total of 46 parents completed the PSE measure, but only 40 were also videotapedplaying with their child. Videos of two parents were lost due to equipment failure.

[2] Because the intergenerational learning program served only 16 families at one time, data for795 this study were gathered over a two-year period and coding was conducted in three phases.

The first coding phase (n = 20 families) was completed independently by one of the authorsand two trained student coders. For the second phase, the same author who coded in phaseone trained three new coders. The new coders practiced rating segments from the first 20tapes so their ratings could be compared not only with each other but also with ratings from

800 the first phase of coding. Once they achieved sufficient agreement, the three codersindependently rated nine additional videotapes. Inter-rater agreement for sensitivity andintrusiveness remained satisfactory, with ICC AQ6= .69 in both cases. Agreement for ratings ofdetachment was somewhat lower for the second group of coders, with ICC = .58. For thethird phase, two authors trained four new coders using the same methods as in phase two.

805 Coders in phase three rated the nine remaining tapes. ICC measures were sensitivity = .73,intrusiveness = .75, and detachment = .70.

[3] To assess whether the quadratic association was robust, we removed outliers for both PSEand sensitivity and reran analyses. For PSE, we removed the parent with the lowest SEPTIscore (164; see Figure 1). Even with this parent removed, the quadratic association between

810 PSE and sensitivity remained significant after controlling for child sex and the linear effect, β= −.34, p = .049. We also reran the analysis after deleting the parent with the lowestsensitivity score (1.93 on a 1–5 scale). Again, even with this case removed, the quadraticassociation between PSE and sensitivity remained significant, β = −.40, p = .019.

References815 Afifi, W. A. (2010). Uncertainty and information management in interpersonal contexts. In S. Smith

& S. R. Wilson (Eds.), New directions in interpersonal communication research (pp. 94–114).Thousand Oaks, CA: Sage.

Ainsworth, M. D. S. (1967). Infancy in Uganda, infant care and the growth of love. Baltimore: JohnHopkins Press.

820 Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. (1978). Patterns of attachment: Apsychological study of the strange situation. Hillsdale, NJ: Lawrence Erlbaum.

Ardelt, M., & Eccles, J. S. (2001). Effects of mothers’ parental efficacy beliefs and promotiveparenting strategies on inner-city youth. Journal of Family Issues, 22, 944–972. doi:10.1177/019251301022008001

825 Bandura, A. (1982). Self-efficacymechanisms in human agency.American Psychologist, 37(2), 122–147.doi:10.1037/0003-066X.37.2.122

Bandura, A. (1986). Social foundations of thought and action: A social-cognitive theory. EnglewoodCliffs, NJ: Prentice-Hall.

Bandura, A. (1989). Regulation of cognitive processes through perceived self-efficacy. Develop-830 mental Psychology, 25, 729–735. doi:10.1037/0012-1649.25.5.729

Barnett, M. A. (2008). Mother and grandmother parenting in low-income three-generational ruralhouseholds. Journal of Marriage and the Family, 70, 1241–1257. doi:10.1111/j.1741-3737.2008.00563.x

Bogenschneider, K., Small, S., & Tsay, J. (1997). Child, parent, and contextual influences on835 perceiving parenting competence among parents of adolescents. Journal of Marriage and the

Family, 59, 345–362. doi:10.2307/353475Bohlin, G., & Hagekull, B. (1987). “Good mothering”: Maternal attitudes and mother-infant

interaction. Infant Mental Health Journal, 8, 352–363. doi:10.1002/1097-0355(198724)8:4%3C352::AID-IMHJ2280080404%3E3.0.CO;2-R

Parental Self-efficacy and Sensitivity 21

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:47

840 Borrego, J., Timmer, S. G., Urquiza, A. J., & Follett, W. C. (2004). Physically abusive mothers’responses following episodes of child noncompliance and compliance. Journal of Consultingand Clinical Psychology, 72, 897–903. doi:10.1037/0022-006X.72.5.897

Bowlby, J. (1969/1982). Attachment and loss: Vol. 1: Attachment. (2nd ed.). New York, NY: BasicBooks (Original work published 1969).

845 Bugental, D. B., & Happaney, K. (2000). Parent–child interaction as a power contest. Journal ofApplied Developmental Psychology, 21, 267–282. doi:10.1016/S0193-3973(99)00038-6

Coleman, P. K., & Karraker, K. H. (1997). Self-efficacy and parenting quality: Findings and futureapplications. Developmental Review, 18(1), 47–85. doi:10.1006/drev.1997.0448

Coleman, P. K., & Karraker, K. H. (2003). Maternal self-efficacy beliefs, competence in parenting, and850 toddlers’ behavior and developmental status. Infant Mental Health Journal, 24(2), 126–148.

doi:10.1002/imhj.10048Conrad, B., Gross, D., Fogg, L., & Ruchala, P. (1992). Maternal confidence, knowledge, and quality

of mother-toddler interactions: A preliminary study. Infant Mental Health Journal, 13,353–362.

855 Corapci, F., & Wachs, T. D. (2002). Does parental mood or efficacy mediate the influence ofenvironmental chaos upon parenting behavior? Merrill-Palmer Quarterly, 48(2), 182–201.doi:10.1353/mpq.2002.0006

Cox, M. J., & Crnic, K. (2003). Qualitative ratings for parent-child interaction at 3–15 months ofage. Unpublished manuscript. Chapel Hill, NC: University of North Carolina.

860 De Wolff, M. S., & van IJzendoorn, M. H. (1997). Sensitivity and attachment: A meta-analysis ofparental antecedents of infant attachment. Child Development, 68, 571–591.

Frosch, C. A., Cox, M. A., & Goldman, B. D. (2001). Infant-parent attachment and parental andchild behavior during parent-toddler storybook interaction. Merrill-Palmer Quarterly, 47,445–474. doi:10.1353/mpq.2001.0022

865 Gross, D., Fogg, L., & Tucker, S. (1995). The efficacy of parent training for promoting positiveparent-toddler relationships. Research in Nursing & Health, 18, 489–499. doi:10.1002/nur.4770180605

Hart, C. H., Newell, L. D., & Olsen, S. F. (2003). Parenting skills and social-communicativecompetence in childhood. In J. O. Greene & B. R. Burleson (Eds.), Handbook of

870 communication and social interaction skills (pp. 753–799). Mahwah, NJ: Lawrence Erlbaum.Izzo, C., Weiss, L., Shanahan, T., & Rodriguez-Brown, F. (2000). Parental self-efficacy and social

support as predictors of parenting practices and children’s socioemotional adjustment inMexican immigrant families. Journal of Prevention and Intervention in the Community, 20,197–213. doi:10.1300/J005v20n01_13

875 Johnston, C., & Mash, E. J. (1989). A measure of parenting satisfaction and efficacy. Journal ofClinical Child Psychology, 18(2), 167–175. doi:10.1207/s15374424jccp1802_8

Jones, T. L., & Prinz, R. J. (2005). Potential roles of parental self-efficacy in parent and childadjustment: A review. Clinical Psychology Review, 25, 341–363. doi:10.1016/j.cpr.2004.12.004

Kaminski, J. W., Valle, L. A., Filene, J. H., & Boyle, C. L. (2008). A meta-analytic review of880 components associated with parent training program effectiveness. Journal of Abnormal

Child Psychology, 36, 567–589. doi:10.1007/s10802-007-9201-9Maloney, E. K., Lapinski, M. K., & Witte, K. (2011). Fear appeals and persuasion: A review and

update of the Extended Parallel Process Model. Social and Personality Psychology Compass,5, 206–219. doi:10.1111/j.1751-9004.2011.00341.x

885 Munz, E., & Wilson, S. R. (in press). Caregiver confirmation and children’s attachment securityduring the transition to kindergarten. Communication Research. AQ7

Sacks, H., Schegloff, E. A., & Jefferson, G. (1974). A simplest systematics for the organization ofturn-taking for conversation. Language, 50, 696–735. doi:10.2307/412243

Semple, D. L, Mash, E. J., Ninowski, J. E., & Benzies, K. M. (2011). The relation between maternal890 symptoms of Attention-Deficit/Hyperactivity Disorder and mother–infant interaction.

Journal of Child and Family Studies, 20, 460–472. doi:10.1007/s10826-010-9413-4Shumow, L., & Lomax, R. (2002). Parental self-efficacy: Predictor of parenting behavior and

adolescent outcomes. Parenting Science and Practice, 2, 127–150. doi:10.1207/S15327922PAR0202_03

22 S.R. Wilson et al.

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:47

895 Teti, D. M., & Gelfand, D. M. (1991). Behavioral competence among mothers of infants in the firstyear: The meditational role of maternal self-efficacy. Child Development, 62, 918–929.doi:10.2307/1131143

Trees, A. R. (2006). Attachment theory: The reciprocal relationship between family communicationand attachment patterns. In D. O. Braithwaite & L. A. Baxter (Eds.), Engaging theories in

900 family communication: Multiple perspectives (pp. 165–180). Thousand Oaks, CA: Sage.Webster-Stratton, C., & Hammond, M. (1997). Treating Children with Early-onset conduct

problems: A comparison of child and parent training interventions. Journal of Consultingand Clinical Psychology, 65(1), 93–109. AQ8doi:10.1037/0022-006X.65.1.93

Wilson, S. R., Rack, J. J., Shi, X., & Norris, A. M. (2008). Comparing physically abusive, neglectful,905 and non-maltreating parents during interactions with their children: A meta-analysis of

observational studies. Child Abuse & Neglect, 32, 897–911. doi:10.1016/j.chiabu.2008.01.003

Parental Self-efficacy and Sensitivity 23

{RJAC}Articles/RJAC911937/RJAC_A_911937_O.3d 15th April 2014 17:47:47