Appetite 91 (2015) 107–113

Contents lists available at ScienceDirect

Appetite j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / a p p e t

Research report

Confirmatory factor analysis of the Feeding Emotions Scale. A measure of parent emotions in the context of feeding ☆ Leslie Frankel a, Jennifer O. Fisher b, Thomas G. Power c, Tzu-An Chen d, Matthew B. Cross e, Sheryl O. Hughes d,* a

Department of Psychological, Health and Learning Sciences, University of Houston, Houston, TX 77004, USA Department of Public Health, Temple University, Philadelphia, PA 19140, USA c Department of Human Development, Washington State University, Pullman, WA 99164, USA d USDA/ARS Children’s Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine, Houston, TX 77030, USA e Department of Health and Human Performance, Texas Obesity Research Center, University of Houston, Houston, TX 77204-6015, USA b

A R T I C L E

I N F O

Article history: Received 14 November 2014 Received in revised form 31 March 2015 Accepted 4 April 2015 Available online 9 April 2015 Keywords: Parent–child dyad Feeding emotions scale Measures Factor analysis Differential emotions scale

A B S T R A C T

Assessing parent affect is important because studies examining the parent–child dyad have shown that parent affect has a profound impact on parent–child interactions and related outcomes. Although some measures that assess general affect during daily lives exist, to date there are only few tools that assess parent affect in the context of feeding. The aim of this study was to develop an instrument to measure parent affect specific to the feeding context and determine its validity and reliability. A brief instrument consisting of 20 items was developed that specifically asks how parents feel during the feeding process. This brief instrument draws on the structure of a well-validated general affect measure. A total of 296 Hispanic and Black Head Start parents of preschoolers completed the Feeding Emotions Scale along with other parent-report measures as part of a larger study designed to better understand feeding interactions during the dinner meal. Confirmatory factor analysis supported a two-factor model with independent subscales of positive affect and negative affect (Cronbach’s alphas of 0.85 and 0.84, respectively). Concurrent and convergent construct validity was evaluated by correlating the subscales of the Feeding Emotions Scale with positive emotionality and negative emotionality from the Differential Emotions Scale – a measure of general adult emotions. Concurrent and convergent criterion validity was evaluated by testing mean differences in affect across parent feeding styles using ANOVA. A significant difference was found across maternal weight status for positive feeding affect. The resulting validated measure can be used to assess parent affect in studies of feeding to better understand how interactions during feeding may impact the development of child eating behaviors and possibly weight status. © 2015 Elsevier Ltd. All rights reserved.

Introduction

Abbreviations: ANOVA, Analysis of Variance between Groups Data Entry; CFI, Comparative Fit Index; CFSQ, Caregiver’s Feeding Styles Questionnaire; CITC, Corrected Item-Total Correlations; DES, Differential Emotions Scale; FES, Feeding Emotions Scale; PANAS, Positive and Negative Affect Schedule; RMSEA, Root Mean Square Error of Approximation; SRMR, Standardized Root Mean Squared Residual; WLSMV, Meanand Variance-adjusted Weighted Least Square. ☆ Acknowledgements: This research was supported by funds from the U. S. Department of Agriculture, Grant No. 2006-55215-16695. This work is a publication of the United States Department of Agriculture (USDA/ARS) Children’s Nutrition Research Center, Department of Pediatrics, Baylor College of Medicine, Houston, Texas, and had been funded, in part, with federal funds from the USDA/ARS under Cooperative Agreement No. 6250-51000-053. The contents of this publication do not necessarily reflect the views or policies of the USDA, nor does mention of trade names, commercial products, or organizations imply endorsement from the U.S. government. * Corresponding author. E-mail address: [email protected] (S.O. Hughes). http://dx.doi.org/10.1016/j.appet.2015.04.008 0195-6663/© 2015 Elsevier Ltd. All rights reserved.

The family feeding environment remains one of the most important contexts in which children’s eating behaviors are socialized. This area of research is essential to understanding the development of childhood obesity because parents lay the foundation for the development of appropriate child eating behaviors (Birch & Fisher, 1998; Sleddens, Gerards, Thijs, de Vries, & Kremers, 2011). Some studies have linked the broader family feeding environment (as measured by feeding styles) to child eating behaviors and child weight status (Hoerr et al., 2009; Hughes, Power, Orlet Fisher, Mueller, & Nicklas, 2005; Hughes et al., 2011; Hughes, Shewchuk, Baskin, Nicklas, & Qu, 2008). Different from goal-directed feeding behaviors (e.g., pressure to eat, restriction), feeding styles incorporate the emotional climate parents create with their children during feeding episodes (Hughes et al., 2005). The emotional climate created by parents includes their tone and affect which plays an important part in how children react to parenting directives around feeding (Hughes

108

L. Frankel et al./Appetite 91 (2015) 107–113

et al., 2011). Although feeding style typologies (authoritarian, authoritative, indulgent, and uninvolved) are defined, in part, by parent tone of voice and affect, these qualities are typically inferred and not directly measured (Hughes et al., 2005, 2011). Directly assessing parent affect is important because studies examining parent– child interactions have shown that parent affect has a profound impact on children’s emotional well-being (Teti, Gelfand, Messinger, & Isabella, 1995). Direct observation of parent–child interactions in the family feeding environment provides insight into the role of parental emotions during feeding. For example, Fiese, Winter, and Botti (2011) and Fiese, Hammons, and Grigsby-Toussaint (2012) observed family mealtimes to examine the relationship between several dimensions of parent–child interactions and child health. Family communication was a key component in their observational coding system. This communication component included interpersonal communication thought to promote appropriate interactions as well as critical communication thought to impair interactions. Fiese et al. (2011) defined critical communication as “disapproving or harsh statements (verbal content and/or tone) regarding another participant’s behavior or personality” (p. 138). These researchers found that families who engage in more critical communication during family meals had children who were more likely to be overweight or obese (Fiese et al 2012). Although these studies of family meals provide important information beyond what parents are saying to their children during feeding episodes including their tone of voice, parent affect was not included as a variable in the coding system (Fiese et al., 2012). A measure of parent affect could inform the emotional interpretation of parent and child interactions during feeding. For example, specific information on positive affect could tell us about parents’ positive regard, warmth, and affection toward the child during feeding, which may alter the child’s response to parenting directives (Belsky, Crnic, & Woodworth, 1995). More importantly, it would provide rich information on the emotional climate of the parent–child feeding dynamic informing prevention studies aimed at reducing childhood obesity. Positive parent affect has been associated with better socialemotional and behavioral outcomes while negative parent affect has been associated with poorer child outcomes (Martin, Clements, & Crnic, 2002). Positive emotional interactions during family meals have been linked to better child health outcomes (Fiese et al., 2012; Jacobs & Fiese, 2007). When parent tone and affect have been directly measured, differences in feeding styles have been found. Hughes et al. (2011) found differences in parent negative (but not positive) affect across feeding styles. This could reflect the fact that in this study parent affect was observed and not self-reported. Most observational studies are relatively underpowered to detect differences. A short validated parent-report measure of parent affect may provide researchers with an easy way to examine affect in larger samples and thus provide better information regarding how affect at mealtimes impacts the eating behaviors of young children. In reviewing the literature, only six studies directly examined parent affect, feeding, and child eating behaviors together in the same study and, in most of those studies, only parent negative affect was measured. In these studies, parent negative affect was associated with a number of socio-environmental influences on children’s eating. Hampson, Tonstad, Irgens, Meltzer, and Vollrath (2010) found that higher maternal negative affectivity was associated with a higher likelihood of providing children under the age of 6 months with sweet drinks and solid foods. Hughes, Shewchuk, Baskin, Nicklas, and Qu (2008) and Hughes et al. (2011) found parental negative affect to be associated with highly controlling feeding of preschoolers. Hughes and colleagues, in a separate study, also found parental negative affect to be associated with an increased perception of feeding problems (Hughes & Shewchuk, 2012). In a longitudinal study, Hafstad, Abebe, Torgersen, and von Soest (2013), found that

maternal negative affectivity when the child was age 1 ½ predicted picky eating at ages 2 ½ and 4 ½. In contrast, child emotional eating was negatively associated with family open expression of affection and emotion in children ages 6 to 8 (Topham et al., 2011). Other studies have examined associations between maternal mental health (e.g., depression) and feeding/eating behaviors but have not explicitly assessed maternal affect (see El-Behadli, Sharp, Hughes, Obasi, & Nicklas, 2015, for a review). A measure of parent affect specifically assessing affect in the feeding domain is needed to better understand the parent–child feeding dynamic. Furthermore, the studies mentioned above assessed general parental affect. Although relevant, assessing parent affect in the context of feeding would provide more specificity to parent–child feeding interactions and enable researchers to tease apart the impact of positive versus negative affect on various types of feeding directives and determine how they influence child eating behaviors. The aim of this study was to develop an instrument to measure parent affect specific to the feeding context. Examination of affect and emotions specific to feeding can provide valuable information on how parent emotions in the context of feeding are related to outcomes such as children’s eating behaviors, children’s food preferences, children’s enjoyment of eating, and quality of family mealtimes. In order to accomplish this, we developed a brief instrument of parental affect during child feeding that draws upon the structure of a well-validated general affect measure (Watson & Clark, 1992; Watson, Clark, & Tellegen, 1988). Researchers have attempted to directly observe parent affect and other parenting behaviors during observations of the family-eating environment but observational studies of family meal environments tend to be small and expensive to conduct (Kong et al., 2013). A novel parentreport questionnaire measuring parent emotion during feeding is less time-intensive and more cost effective than traditional observational studies and could advance our understanding of the emotional climate of parent–child feeding interactions. Such a measure would also contribute to smaller, observational studies, by providing additional data on affect from the parent’s perspective. Initial evidence of concurrent and convergent construct validity was assessed by evaluating associations of parents’ general and feedingspecific emotional experience, and evidence of concurrent and convergent criterion validity was assessed by evaluating associations of parent affect with feeding styles. Subjects and methods Participants Participants were 296 Hispanic and Black parents of preschoolaged children recruited from Head Start centers in a large urban city in southwestern United States. The primary caregiver (referred to as parent hereafter) was defined as the person who was most often responsible for what the child ate outside of the Head Start school day. The families were part of a larger study designed to observe parent–child interactions during the dinner meal. Exclusion criteria included severe food allergies and known developmental delays (e.g., autism). Parents completed questionnaires as part of this larger study – these questionnaires were chosen because they assessed constructs that may impact the child eating environment. Characteristics of the participants are shown in Table 1. Procedures Parents were recruited for the study during child drop off and pick up at the Head Start centers, through presentations at Head Start meetings, and flyers posted at the Head Start centers. This study was part of a larger study involving observations in participants’

L. Frankel et al./Appetite 91 (2015) 107–113

Table 1 Demographics and lifestyle characteristics of participants (n = 296). Demographic variables Child Age, years [M(SD)] Gender (%) Male Female Child BMI percentile Normal Overweight Obese Parent Age, years [M(SD)] Parent BMI [M(SD)] Parent BMI categories (%) Normal Overweight Obese Race/Ethnicity (%) Black Hispanic Marital status Married Maternal education At least high school diploma Feeding styles Authoritative Authoritarian Indulgent Uninvolved

% 4.42 (0.71) 51 49 62.5 18.6 18.9 31.75 (7.80) 31.72 (7.85) 18.6 28.0 53.4 46.8 53.2 45.8 55.2 20 27 31 22

homes. Packets of questionnaires to be completed by parents were provided at the end of each home visit and completed questionnaires were returned to staff members at the subsequent home visit. Questionnaires were available in both English and Spanish. Spanish versions were translated into Spanish and back translated into English by Spanish-speaking staff members. No issues in understanding the measures were reported by participants. Graduated incentives were provided to the families at the end of each home visit. Consent for study participation was obtained at the beginning of the first home visit and confidentiality was assured. The Institutional Review Board of Baylor College of Medicine approved the study. Measures Feeding emotions scale (FES) The FES (available upon request from the corresponding author) assesses two broad aspects of parents’ emotions while feeding their children, positive and negative affect, that have been previously identified as “dominant dimensions of emotional experience” (Watson & Clark, 1992, p. 443). The format of the FES directly follows the format of the Positive and Negative Affect Schedule (PANAS), a wellvalidated general measure of affect (Crawford & Henry, 2004; Watson & Clark, 1992). The PANAS assesses the extent to which individuals experience a series of affective words and phrases during a specified period (positive examples: interested and strong; negative examples: jittery and afraid) (Watson et al., 1988). Drawing upon this format, the authors (SOH, JOF) chose 20 affective terms reflecting emotions that parents may experience during feeding episodes. These words were chosen to be specific to the context of feeding so they differ from the PANAS items in that they are more descriptive of emotions around feeding. Positive affect descriptors included energetic, loved, content, rewarded, relaxed, reassured, connected, and happy. Negative affect descriptors included rejected, unappreciated, unloved, anxious, panicky, exhausted, lonely, tense, angry, frustrated, strung out, and overwhelmed. Parents were asked how often they typically experience these emotions during feeding

109

their preschooler on a 5-point Likert scale (1 [never], 2 [rarely], 3 [sometimes], 4 [most of the time] and 5 [always]) and given the instructions, “Below is a list of words that describe different feelings and emotions. Circle the best answer. If you are not certain, make your best guess. How often have you felt this way when FEEDING your child?” Differential emotions scale (DES) The DES is a 36-item self-report questionnaire that measures general emotions in daily life on a 5-point scale ranging from rarely or never to very often (Izard, Libero, Putnam, & Haynes, 1993). The positive emotionality composite from the DES reflects three discrete subscales (joy, interest and surprise). The negative emotionality composite reflects nine discrete subscales (inner-directed hostility, shyness, disgust, shame, anger, sadness, contempt, guilt and fear). Reliability and validity have been established for the DES (Boyle, 1984). Caregiver’s feeding styles questionnaire (CFSQ) The Caregiver’s Feeding Styles Questionnaire (CFSQ) was used to assess parent feeding styles (Hughes et al., 2005). The CFSQ is a self-report instrument that assesses parent feeding style in a similar way to general parenting styles (Maccoby & Martin, 1983) (see Hughes et al., 2005, for more details). Four feeding style typologies are based on levels of parent demandingness and responsiveness in feeding their children: authoritative (high demandingness, high responsiveness), authoritarian (high demandingness, low responsiveness), uninvolved (low demandingness, low responsiveness), and indulgent (low demandingness, high responsiveness). Responsiveness and demandingness are calculated through child-centered feeding items (those that promote child autonomy such as reasoning, complimenting, and helping the child to eat) and parentcentered feeding items (those that control child eating through external pressure such as forcing and reward contingencies). The CFSQ can be scored using the typological approach as described above or the dimensional approach (continuous scores) to test alternate hypotheses. The continuous scores are derived to determine specific aspects of the feeding domain including parent-centered/ high control, parent-centered/contingency management, and childcentered feeding practices (Hughes et al., 2006). Evidence of internal consistency and test–retest reliability as well as convergent and predictive validity has been demonstrated (Hennessy, Hughes, Goldberg, Hyatt, & Economos, 2010; Hughes et al., 2005, 2008). Anthropometrics Height and weight were measured on the parent and child by trained staff with height measured to the nearest 0.1 cm and weight to the nearest 0.1 kg (Lohman, Roche, & Martorell, 1988). Parents and children were asked to dress in light clothing and remove their shoes. Measurements were taken twice and averaged. If a difference of more than .5 cm for height or 3 kg for weight was seen, a third measurement was taken and the three measures were averaged. Parent BMI was calculated based on CDC criteria and classified on CDC cutoffs as healthy weight (BMI < 24.9), overweight (BMI > 25) or obese (BMI > 30). Height and weight scores on the children were converted to age- and gender-specific BMI z-scores using the revised 2000 growth charts from the Centers for Disease Control and Prevention (Kuczmarski et al., 2002). Statistical analysis Means, standard deviations, and corrected item-total correlations (CITC) for each item were calculated. The internal consistency reliability of each subscale was examined using Cronbach’s alpha. Values greater than 0.3 and 0.7 were considered acceptable for CITC and internal consistency reliability, respectively (Nunnally, 2010).

110

L. Frankel et al./Appetite 91 (2015) 107–113

A confirmatory factor analysis (CFA) of the two-factor model was run in Mplus version 7.11 with the Mean- and Variance-adjusted Weighted Least Square (WLSMV) which is appropriate for the evaluation of ordered categorical data (Muthén & Muthén, 1998–2011) to confirm the subscales. Positive and negative affect were conceptualized as two correlated factors. The chi-square statistic was not used to determine model fit because of its assumption of normality and sensitivity to sample size (Fan, Thompson, & Wang, 1999). The Two-Index Presentation Strategy was used to assess model fit (Hu & Bentler, 1999). This presentation strategy states that one of the following is true: a) the Tucker-Lewis Index (TLI) > 0.96 and the Standardized Root Mean Squared Residual (SRMR) < 0.09; b) the Root Mean Square Error of Approximation (RMSEA) < 0.06 and the SRMR < 0.09; or c) the Comparative Fit Index (CFI) > 0.96 and the SRMR < 0.09. Concurrent and convergent construct validity was evaluated by correlating the subscales of the FES with the subscales of a general emotion questionnaire. Concurrent and convergent criterion validity was assessed as follows: 1) by testing mean differences in affect across feeding styles using Analysis of Variance (ANOVA) and Tukey’s post hoc tests; and 2) by correlating the subscales of the FES with the parent-centered/high control and child-centered scales of the CFSQ. In addition, analysis of differences in positive and negative FES between maternal weight statuses was conducted using ANOVA. A significance level of P < .05 was used for all analyses. Results Means, standard deviations, corrected item-total correlations, and internal consistency Means, standard deviations, and CITC are presented in Table 2. All CITC for individual items were above 0.3, indicating that all items correlated adequately with the rest of the corresponding subscale. Cronbach’s alphas were 0.85 for the positive subscale and 0.84 for the negative subscale demonstrating adequate internal consistency.

follows: RMSEA = 0.07, CFI = 0.96, χ2 (df = 169) = 413.86, P < .001, TLI = 0.95, SRMR = 0.08 (Hu & Bentler, 1999). None of the 20 items was removed from the two-factor model (see Fig. 1). The positive and negative factors were significantly correlated (r = −0.37, P < .01). Concurrent and convergent construct validity The positive affect subscale of the FES and the positive composite subscale of DES were significantly correlated (r = .43, P < .001). Similarly, the negative affect subscale of the FES and the negative composite subscale of the DES were significantly correlated in the expected direction (r = .44, P < .001). Concurrent and convergent criterion validity Positive affect differed across feeding styles, F(3, 293) = 6.78, P < .001. Parents with an uninvolved feeding style reported significantly lower positive affect (M = 28.80; SD = 6.11) compared to authoritative parents (M = 32.53; SD = 5.17) and indulgent parents (M = 32.20; SD = 5.19). Both comparisons were significant at P < .01. Parents with an authoritarian feeding style were not different from the other feeding styles on positive affect (M = 30.21; SD = 5.86). Negative affect did not differ across feeding styles (P > .05). The positive affect scale of the FES and CFSQ child-centered scale were significantly related (r = 0.23, P < .001). The negative affect scale of the FES and the CFSQ parent-centered/high control scale were significantly related (r = 0.13, P < .05). Positive/negative feeding affect and maternal weight status A significant difference was found across maternal weight status for positive feeding affect. Overweight parents reported significantly higher positive feeding affect compared to obese parents (P < .05). The means for positive feeding affect for healthy weight, overweight, and obese parents were 30.98 (SD = 6.37), 32.23 (SD = 4.42), and 30.26 (SD = 6.00), respectively. No significant differences among parent weight status were found for negative feeding affect.

Confirmatory factor analysis Discussion The CFA model with no modifications suggested a good fit of the factor structure according to the two-index criteria with indices as

Table 2 Calculated Means, Standard Deviations, and Corrected Item-Total Correlations for each item and internal consistency for each subscale.

Positive affect (Cronbach’s α = 0.85) Energetic Loved Content Rewarded Relaxed Reassured Connected Happy Negative affect (Cronbach’s α = 0.84) Rejected Unappreciated Unloved Anxious Panicky Exhausted Lonely Tense Angry Frustrated Overwhelmed Strung out

M

SD

CITC

3.71 4.44 4.08 4.05 3.77 3.62 3.27 4.21

0.93 0.83 0.93 0.99 0.96 1.11 1.41 0.90

0.60 0.65 0.65 0.67 0.67 0.57 0.37 0.65

1.28 1.42 1.32 1.91 1.25 2.16 1.46 1.72 1.64 1.74 1.82 1.46

0.63 0.75 0.74 1.17 0.62 1.03 0.78 0.91 0.79 0.87 1.01 0.80

0.45 0.49 0.40 0.30 0.49 0.56 0.52 0.67 0.58 0.67 0.53 0.46

It is well known that emotions that parents exhibit with their children play an important part in shaping child emotional regulation (Eisenberg et al., 1992; Pollak & Thoits, 1989) as well as behavioral outcomes (Eisenberg et al., 1999). As such, parent affect plays an important role in parent–child interactions. However, little attention has been given on examining parent emotions in the feeding context. The aim of this study was to develop an instrument designed to measure how parents feel during eating episodes with their children. A 20-item brief measure was developed and validated with low-income Head Start parents of preschoolers. Confirmatory factor analysis supported a two-factor model with independent subscales of positive affect and negative affect. Concurrent and convergent construct validity was supported by associations with a well-validated general measure of affect (Watson & Clark, 1992; Watson et al., 1988), and evidence of concurrent and convergent criterion validity was supported by differences in feeding emotions across parent feeding styles. Internal consistency was established through Cronbach alphas. This instrument – The Feeding Emotions Scale – is an important contribution to the childhood obesity literature because it extends the measurement of parent– child feeding interactions by providing a specific measure of parent affect and emotion in this context that is vitally important to the understanding of parents’ roles in the development of child eating behaviors. In addition, the instrument was developed with lowincome ethnically diverse families who are at a higher risk for developing childhood obesity (Ogden, Carroll, Kit, & Flegal, 2014).

L. Frankel et al./Appetite 91 (2015) 107–113

111

Energetic 0.74 0.84 0.80 0.78

Positive Affect

0.80

Loved Content Rewarded Relaxed

0.72 0.43 0.81

Reassured Connected Happy

Rejected 0.66 0.68 0.65 0.38

Unappreciated Unloved Anxious

0.71

Negative Affect

Panicky 0.71 0.77 0.83 0.73 0.85

Note: Factor loadings are standardized. *All loadings were significant at the level of 0.001

0.63 0.60

Exhausted Lonely Tense Angry Frustrated Overwhelmed Strung Out

Fig. 1. Graphical representation of the Confirmatory Factor Analysis of the 20-item Feeding Emotions Scale.

Although the framework of the Feeding Emotions Scale is similar to the PANAS, the specific items differ from the PANAS items. The feeding context of this measure lends itself to different emotions than the more general PANAS. The Feeding Emotions Scale items were chosen because the scale developers (SOH, JOF) believed these items to be descriptive of emotions around feeding. Researchers who are interested in parent affect during feeding may include this measure in their studies to gain a deeper understanding of parent emotion and affect during feeding. Although our analysis revealed associations between the positive and negative subscales of the FES and general parent affect, these associations were considered moderate. This finding was expected because the associations were between feeding-specific affect and parents’ affect in a more general context (Cunningham, Preacher, & Banaji, 2001). Costanzo and Woody (1985) posited that parenting is domain-specific. Researchers believe that measures developed around feeding may be more closely related to children’s eating behavior and health outcomes and may prove to be more predictive of childhood obesity (Hughes et al., 2008).

In the current study, we found differences across the feeding styles on positive affect but not negative affect which is the opposite of what Hughes and colleagues found in a previous study (Hughes et al., 2011). In the Hughes et al. (2011) study, affect was observed whereas the current study uses self-report. One possibility is that parents are less likely to report as many negative emotions in their interactions with their children as are typically seen in observational research. To this point, the means for negative affect reported by parents were lower than the means reported by parents for positive affect. Alternatively, parents may not perceive their own emotions as negative and therefore not report them as such even though these emotions may be observed by others as negative. Further studies should be conducted to tease apart why there are discrepant findings across the two studies. In the current study, we found that the positive affect subscale of the Feeding Emotions Scale correlated significantly with the CFSQ child-centered items and the negative affect subscale of the Feeding Emotions Scale correlated with the CFSQ parent-centered items. These findings make theoretical sense given that the child-centered

112

L. Frankel et al./Appetite 91 (2015) 107–113

items in the CFSQ tend to be more positive in nature (e.g., praising the child) and CFSQ parent-centered items tend to be more negative in nature (e.g., threatening the child) (Hughes et al., 2006). Significant differences across maternal weight status for positive affect in the context of feeding were also found with overweight parents reporting higher positive affect in the context of feeding than obese parents. This finding is consistent with another study that reported significantly higher general positive affect in overweight women than obese women (Jorm et al., 2003). One possible explanation for this finding is that overweight mothers may be more food responsive, in general, thus enjoying the overall eating experience including feeding their children. However, obese mothers may experience more anxiety and concern over eating and find feeding less positive because of concerns about her own and her child’s weight status. Similarly, healthy weight mothers may be less food responsive than overweight mothers finding the eating and feeding experience just one of many daily routines in her busy life. In this study, we conducted a confirmatory factor analysis of the two-factor model of the Feeding Emotions Scale – a scale developed to measure parent emotions in the feeding context. Adequate model fit of the Feeding Emotions Scale was demonstrated with an ethnically diverse sample and evidence of validity was supported through associations with general parent emotions and feeding styles. Researchers can easily use the Feeding Emotions Scale to assess affect in large groups of parents and successfully measure parent emotion in the context of feeding. This questionnaire will aid researchers in better understanding the dynamics between parents and their children during feeding interactions. Clinicians interested in the family mealtimes may use this instrument to identify and intervene with parents whose mealtime experiences with their children are highly negative. There were several limitations to this study. Test–retest reliability was not conducted on this measure. Furthermore, data were not collected on fathers in the families in this low-income sample. Future studies should include information on fathers in the feeding context to examine how they feel during feeding episodes. Given that this sample was all Hispanic and Black it is important for future studies to investigate how this measure performs with samples of different racial/ethnic backgrounds. We did not find differences in negative affect across parent feeding styles, but it would be interesting to see if differences in negative affect across parent feeding styles were found in samples of varying racial/ethnic backgrounds. Test–retest data need to be collected to evaluate reliability. Finally, direct observations of parental affect during feeding should be included in future research to further evaluate validity.

Conclusions A growing body of literature supports the importance of parent affect and emotions in studying parent–child interactions and their impact on child outcomes. The Feeding Emotions Scale provides an easy way to measure parent affect during feeding episodes with children. The measurement of this construct may be useful in studies examining the development of child eating behaviors and their impact on childhood obesity.

References Belsky, J., Crnic, K., & Woodworth, S. (1995). Personality and parenting. Exploring the mediating role of transient mood and daily hassles. Journal of Personality, 63, 905–929. Birch, L. L., & Fisher, J. O. (1998). Development of eating behaviors among children and adolescents. Pediatrics, 101(3 Pt. 2), 539–549. Boyle, G. J. (1984). Reliability and validity of Izard’s Differential Emotions Scale. Personality and Individual Differences, 5(6), 747–750.

Costanzo, P. R., & Woody, E. Z. (1985). Domain-specific parenting styles and their impact on the child’s development of particular deviance. The example of obesity proneness. Journal of Social and Clinical Psychology, 3(4), 425–445. Crawford, J. R., & Henry, J. D. (2004). The positive and negative affect schedule (PANAS). Construct validity, measurement properties and normative data in a large non-clinical sample. British Journal of Clinical Psychology, 43(Pt. 3), 245– 265. Cunningham, W. A., Preacher, K. J., & Banaji, M. R. (2001). Implicit attitude measures. Consistency, stability, and convergent validity. Psychological Science, 12(2), 163–170. Eisenberg, N., Fabes, R. A., Carlo, G., Troyer, D., Speer, A. L., Karbon, M., et al. (1992). The relations of maternal practices and characteristics to children’s vicarious emotional responsiveness. Child Development, 63(3), 583–602. Eisenberg, N., Fabes, R. A., Shepard, S. A., Guthrie, I. K., Murphy, B. C., & Reiser, M. (1999). Parental reactions to children’s negative emotions. Longitudinal relations to quality of children’s social functioning. Child Development, 70, 513–534. El-Behadli, A. F., Sharp, C., Hughes, S. O., Obasi, E. M., & Nicklas, T. A. (2015). Maternal depression, life stress, and feeding styles. Toward a framework for theory and research in child obesity. British Journal of Nutrition, 113, S55–S71. Fan, X., Thompson, B., & Wang, L. (1999). Effects of sample size, estimation methods, and model specification on structural equation modeling fit indexes. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 56–83. Fiese, B. H., Hammons, A., & Grigsby-Toussaint, D. (2012). Family mealtimes. A contextual approach to understanding childhood obesity. Economics & Human Biology, 10(4), 365–374. Fiese, B. H., Winter, M. A., & Botti, J. C. (2011). The ABC’s of family mealtimes. Observational lessons for promoting healthy outcomes for children with persistent asthma. Child Development, 82(1), 133–145. Hafstad, G. S., Abebe, D. S., Torgersen, L., & von Soest, T. (2013). Picky eating in preschool children. The predictive role of the child’s temperament and the mother’s negative affectivity. Eating Behaviors, 14, 274–277. Hampson, S. E., Tonstad, S., Irgens, L. M., Meltzer, H. M., & Vollrath, M. E. (2010). Mothers’ negative affectivity during pregnancy and food choices for their infants. International Journal of Obesity (Lond), 34, 327–331. Hennessy, E., Hughes, S. O., Goldberg, J. P., Hyatt, R. R., & Economos, C. D. (2010). Parent behavior and child weight status among a diverse sample of underserved rural families. Appetite, 54, 369–377. Hoerr, S. L., Hughes, S. O., Fisher, J. O., Nicklas, T. A., Liu, Y., & Shewchuk, R. M. (2009). Associations among parental feeding styles and children’s food intake in families with limited incomes. International Journal of Behavioral Nutrition and Physical Activity, 6, 55. Hu, L.-T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis. Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6(1), 1–55. Hughes, S. O., Anderson, C. B., Power, T. G., Micheli, N., Jaramillo, S., & Nicklas, T. A. (2006). Measuring feeding in low-income African-American and Hispanic parents. Appetite, 46, 215–223. Hughes, S. O., Power, T. G., Orlet Fisher, J., Mueller, S., & Nicklas, T. A. (2005). Revisiting a neglected construct. Parenting styles in a child-feeding context. Appetite, 44, 83–92. Hughes, S. O., Power, T. G., Papaioannou, M. A., Cross, M. B., Nicklas, T. A., Hall, S. K., et al. (2011). Emotional climate, feeding practices, and feeding styles. An observational analysis of the dinner meal in Head Start families. International Journal of Behavioral Nutrition and Physical Activity, 8, 60. Hughes, S. O., & Shewchuk, R. M. (2012). Child temperament, parent emotions, and perceptions of the child’s feeding experience. International Journal of Behavioral Nutrition and Physical Activity, 9, 64. Hughes, S. O., Shewchuk, R. M., Baskin, M. L., Nicklas, T. A., & Qu, H. (2008). Indulgent feeding style and children’s weight status in preschool. Journal of Developmental & Behavioral Pediatrics, 29, 403–410. Izard, C. E., Libero, D. Z., Putnam, P., & Haynes, O. M. (1993). Stability of emotional experiences and their relations to traits of personality. Journal of Personality and Social Psychology, 64, 847–860. Jacobs, M. P., & Fiese, B. H. (2007). Family mealtime interactions and overweight children with asthma. Potential for compounded risks? Journal of Pediatric Psychology, 32(1), 64–68. Jorm, A. F., Korten, A. E., Christensen, H., Jacomb, P. A., Rodgers, B., & Parslow, R. A. (2003). Association of obesity with anxiety, depression and emotional well-being. A community survey. Australian and New Zealand Journal of Public Health, 27(4), 434–440. Kong, A., Jones, B. L., Fiese, B. H., Schiffer, L. A., Odoms-Young, A. M., Kim, Y., et al. (2013). Mealtime interactions in racially/ethnically diverse families with preshool-age children. Eating Behaviors, 14, 451–455. Kuczmarski, R., Ogden, C., Guo, S., Grummer-Strawn, L. M., Flegal, K. M., Mei, Z., et al. (2002). CDC growth charts for the United States. Methods and development. Vital Health Statistics, 11(246), 1–190. Lohman, T. G., Roche, A. F., & Martorell, R. (1988). Anthropometric standardization reference manual. Champaign, IL: Human Kinetics. Maccoby, E., & Martin, J. (1983). Socialization in the context of the family. Parent-child interaction. In P. H. Mussen (Ed.), Handbook of child psychology (pp. 1–101). New York, NY: Wiley. Martin, S. E., Clements, M. L., & Crnic, K. A. (2002). Maternal emotions during mother-toddler interaction. Parenting in affective context. Parenting: Science and Practice, 2(2), 105–126. Muthén, L. K., & Muthén, B. O. (1998–2011). Mplus (version 7.11). [computer software].Los Angeles, CA: Muthén & Muthén.

L. Frankel et al./Appetite 91 (2015) 107–113

Nunnally, J. C. (2010). Psychometric theory 3E. Noida: Tata McGraw-Hill Education. Ogden, C. L., Carroll, M. D., Kit, B. K., & Flegal, K. M. (2014). Prevalence of childhood and adult obesity in the United States, 2011–2012. Journal of the American Medical Association, 311(8), 806–814. Pollak, L. H., & Thoits, P. A. (1989). Processes in emotional socialization. Social Psychology Quarterly, 52(1), 22–34. Sleddens, E. F. C., Gerards, S. M. P. L., Thijs, C., de Vries, N. K., & Kremers, S. P. J. (2011). General parenting, childhood overweight and obesity-inducing behaviors. A review. International Journal of Pediatric Obesity, 6, e12–e27. Teti, D. M., Gelfand, D. M., Messinger, D. S., & Isabella, R. (1995). Maternal depression and the quality of early attachment. An examination of infants, preschoolers, and their mothers. Developmental Psychology, 31, 364–376.

113

Topham, G. L., Hubbs-Tait, L., Rutledge, J. M., Page, M. C., Kennedy, T. S., Shriver, L. H., et al. (2011). Parenting styles, parental response to child emotion, and family emotional responsiveness are related to child emotional eating. Appetite, 56, 261–264. Watson, D., & Clark, L. A. (1992). On traits and temperament. General and specific factors of emotional experience and their relations to the Five-Factor Model. Journal of Personality, 60, 441–476. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect. The PANAS Scales. Journal of Personality and Social Psychology, 54, 1063–1070.