How Parents Influence Healthy Eating Behavior Children Age 1 12

Introduction
Children are highly dependent on their parents because they are their sole providers. Parents' primary responsibility is to provide the basic needs - food, shelter and clothing - of their children. Therefore, parents shape the eating habits of children especially those under the age of 12 years. Generally, children are usually ready to learn how to eat new foods. They also observe the eating behavior of adults around them (Reicks, et al.). However, their eating behaviors evolve as they grow old. Numerous studies have identified factors that influence children eating behavior. They include living condition, access to food, number of caretakers or family members nearby, employment status, age, gender and health condition (Savage, et al.). This paper will estimate the effects that the above factors have on the eating habits of children.
Data
The data for this project was compiled from various internet sources. All the statistical analysis was carried out using Microsoft Excel statistical software. Descriptive statistics indicates the mean, median, standard deviation, maximum, and minimum values of each variable. The correlation coefficient, r, measures the strength of the linear relationship between any two variables. Regression analysis predicts the influence of one or more explanatory (independent) variables on the dependent variable
Descriptive Statistics
Descriptive statistics were used to describe the variables used in this project. The results are displayed in Table A1 (Appendix A). Eating behavior scores ranges from 41 to 100 (M = 71.07, SD = 17.47). A higher eating behavior score reflects a healthy eating behavior. The average age of the subjects is 6.39 years. Most of the subjects live in a developed area (60.20%) and their parents are employed (65.31%). Also, most of the households are made up of both parents. Almost half of the subjects were male (53.06%). 54.08% of the subjects do not use electronics at mealtimes. Approximately half of the subjects (56.12%) confirmed that the availability of food was limited.
Correlation
Correlation results are displayed in Table B1 (Appendix B). It is clear that the independent variables are not correlated.
Regressions and Interpretations
Regression analysis was performed to predict eating habits among children. Four different regression equations were estimated. Each of the equations is described below.
Regression Equation 1
Eating Behavior = ?0 + ?1living location + ?2Access to food + ?3Age + ?4Gender + ?5Electronic use
Equation (1) is the base regression model for estimating eating behavior in children. It shows the linear relationship between eating behavior and the key explanatory variables (living location, access to food, age, gender, and electronic use). The Excel results of estimating this equation are displayed in Table C1 (Appendix C). The estimated equation is follows:
Eatingbehavior = 75.598 + 2.253 livloc - 7.643Foodacc - 0.572 Age
(t) (15.11) (0.62) (- 2.20) (- 1.17)
- 4.268Gender + 7.375Elec use R2 = 0.1133
(- 1.23) (2.10)
All variables are insignificant at 5 percent level expect except access to food (p-value = 0.03) and electronic use (p-value = 0.04). I further perform t-test to determine whether access to food has a negative effect on eating behavior and whether the use of electronics during mealtime affects eating behavior. First, the null and alternative hypothesis of access to food is
H0: ?2 = 0
HA: ?2 0
The test statistic for access to food is – 2.20. At 5 percent significance level, the critical value of t-distribution with N – 5 = 93 degrees of freedom is, t (0.95, 93) = 0.063. Since the calculated value falls in the rejection region, I reject the null hypothesis that ?2 = 0 and conclude that the coefficient of access to food is nonzero (Hill, et al. 109).
Secondly, the null and alternative hypothesis of electronic use is
H0: ?5 = 0
HA: ?5 0
Since t = 2.10 is greater than 0.063, I reject the null hypothesis that ?5 = 0 and conclude that the coefficient of electronic use is statistically significant. This test confirms that if children do not use electronics during mealtimes, their eating habits improve.
R- Squared is 0.1133. It means that the regression model explains 11.33% of the variation in eating behavior.
Regression Equation 2
Eating Behavior = ?0 + ?1living location + ?2Access to food + ?3Age + ?4Gender + ?5Electronic use + ?6Household
In equation (2), the first proxy, the household is added to the model. The Excel results of estimating this equation are displayed in Table D1 (Appendix D). The estimated regression equation is as follows:
Eatingbehavior = 75.366 + 2.225 livloc - 7.674Foodacc - 0.572 Age
(t) (14.1) (0.61) (-2.20) (- 1.16)
- 4.282Gender + 7.414Elecuse + 0.470Household R2 = 0.1134
(- 1.22) (- 2.09) (0.14)
In this model, household is statistically insignificant (p-value = 0.892791184). Therefore, household type (single parent or both parents) does not influence the eating behavior of children. The value of R – Squared remained unchanged at 0.1134. It means that the addition of family structure did not improve the fit of the model. The coefficients of the variables changed slightly compared to the coefficients of the base model. The estimate of electronic use during mealtimes increased from 7.375 to 7.414. However, the coefficient of age remained unchanged at – 0.572.
Regression Equation 3
Eating Behavior = ?0 + ?1living location + ?2Access to food + ?3Age + ?4Gender + ?5Electronic use + ?6Employment status
In equation (3), the second proxy, employment status is included in the base model. The Excel results of estimating this equation are displayed in Table E1 (Appendix E). The estimated equation is as follows:
Eatingbehavior = 76.937 + 2.386 livloc - 7.630Foodacc - 0.600 Age
(t) (14.09) (0.65) (- 2.19) (- 1.22)
3.833Gender + 7.449Elec use – 2.307Empstatus R2 = 0.1170
(- 1.08) (2.11) (- 0.62)
The effects of the second proxy (employment status) are almost similar to the effects of the first proxy (household). The coefficient of employment status is insignificant at 5 percent level (p-value is 0.53510793). The value of R- Squared changed slightly. It means that the inclusion of the second proxy does not improve the adequacy of the base model. The coefficients of this model are almost similar to the coefficients of the base model. For example, the estimate of access to food is – 7.630 while in the base model it is – 7.643.
Regression Equation 4
Eating Behavior = ?0 + ?1living location + ?2Access to food + ?3Age + ?4Gender + ?5Electronic use + ?6Employment status + ?7Household
In equation (4), both proxies are included in the base model. The Excel results for estimating this equation are displayed in Table F1 (Appendix F). The estimated regression equation is as follows:
Eatingbehavior = 76.775 + 2.367 livloc - 7.650Foodacc - 0.600 Age – 3.847Gender
(t) (13.21) (0.64) (- 2.18) (- 1.21) (- 1.07)
+ 7.474Elec use – 2.282Empstatus + 0.298Household R2 = 0.1170
(2.10) (- 0.61) (0.09)
Both employment status (p-value = 0.54319) and household (p-value = 0.932424) is statistically insignificant at 5 percent level. Furthermore, including both proxies into the base model does not improve the fit of the model. The value of R-Squared slightly increases from 0.1133 to 0.1170. Overall, both proxies do not influence eating habits in children.
F-Test
Global F-test was calculated to determine the overall significance of the model in predicting eating behavior in children (Hill, et al 223). The null hypothesis (H0) is that all coefficients of the independent variables are equal to zero. The alternative hypothesis (HA) is that at least one of the coefficients is not equal to zero. Therefore, the null and alternative hypothesis for this test is
H0: ?0 = ?1 = ?2 = ?3 = ?4 = ?5 = ?6 = ?7 = 0
HA: At least one of the coefficients is nonzero.
Using ? = 0.05, the critical value from F (1, 90) –distribution is Fc = F (0.95, 1, 90) = 3.947. Therefore, the rejection region of F 3.947. The test statistic is 1.705. Since F = 1.705 is less than Fc = 3.947, I do not reject the null hypothesis that ?0 = ?1 = ?2 = ?3 = ?4 = ?5 = ?6 = ?7 = 0 and conclude that regression equation 4 is inadequate in explaining the variability of children eating behavior (Hill, et al. 225).
To determine the overall significance of the base regression model, I perform the Global F-test. The null and alternative hypothesis is as follows:
H0: ?0 = ?1 = ?2 = ?3 = ?4 = ?5 = 0
HA: At least one of the coefficients is nonzero
Using ? = 0.05, the critical value from F (6, 92) – distribution is Fc = F (0.95, 6, 92) = 2.199. Therefore, the rejection region of F 2.199. The test statistic is 2.350. Since 2.350 is greater than 2.199, I reject the null hypothesis that ?0 = ?1 = ?2 = ?3 = ?4 = ?5 = 0 and conclude that at least one of the coefficients is nonzero (Cooke and Wardle). The F-test indicates that regression equation 1 is adequate in explaining the variability of children eating behavior.
Conclusion
Many factors influence a child's eating behavior. In this project, it is clear that access to food and the use of electronics influences a child eating behavior. There is a negative relationship between access to food and eating behavior. It implies that if food availability is limited, then the eating habits of any given child is unhealthy. A positive relationship exists between electronic use and eating behavior. Other factors such as living location, gender, age, employment status, household type and employment status are statistically insignificant.
The base regression model (regression equation 1) is adequate in explaining the variability in a child's eating habits. Employment status of the parent and the type of household are irrelevant variables. These variables complicate the base model unnecessarily. However, further research should be carried out to determine how these two proxies affect eating behavior.






Works Cited
Cooke, Lucy J., and Jane Wardle. "Age and gender differences in children's food preferences." British Journal of Nutrition, vol. 93, no. 5, 2005, pp. 741-746, DOI: 10.1079/BJN20051389.
Hill, R. C., et al. Principles of Econometrics. 4th ed., Wiley, 2011.
Reicks, Marla, et al. "Influence of Parenting Practices on Eating Behaviors of Early Adolescents during Independent Eating Occasions: Implications for Obesity Prevention." Nutrients, vol. 7, no. 10, 2015, pp. 8783-8801, www.ncbi.nlm.nih.gov/pmc/articles/PMC4632451/.
Savage, Jennifer S., et al. "Parental Influence on Eating Behavior: Conception to Adolescence." The Journal of Law, Medicine & Ethics, vol. 35, no. 1, 2007, pp. 22-34, www.ncbi.nlm.nih.gov/pmc/articles/PMC2531152/.


Appendix A
Table A1: Descriptive Statistics
 Variables
Description
Mean
Median
Standard Deviation
Minimum
Maximum

Eatingbehavior
Eating behavior scores
71.07
72
17.47
41
100

Livloc
Dummy variable = 1 if the subject lives in a developed area, otherwise 0
0.60
1
0.49
0
1

Foodacc
Dummy variable = 1 if healthy food is easily accessible, otherwise 0
0.44
0
0.50
0
1

Age
Actual years of the subject
6.39
6
3.58
1
12

Gender
Dummy variable = 1 if male, 0 if female.
0.53.
1
0.50
0
1

Elecuse
Dummy variable = 1 if electronics made available during meal times, otherwise 0
0.46
0
0.50
0
1

Empstatus
Dummy Variable = 1 if the parent is employed full-time, otherwise 0
0.65
1
0.48
0
1

Household
Dummy Variable = 1 if the household has both parents, otherwise 0
0.54
1
0.50
0
1


Note: Livloc = living location, Foodacc = Access to food, Elecuse = Electronic use, Empstatus = Employment status

Appendix B
Table B1: Correlations
 
Livloc
Foodacc
Age
Gender
Elecuse
Empstatus
Household

Livloc
1







Foodacc
0.0887
1






Age
0.1120
0.0886
1





Gender
0.1125
0.0336
- 0.1331
1




Elecuse
- 0.2130
0.0105
- 0.0429
0.0050
1



Empstatus
0.0644
- 0.0035
- 0.1133
0.2165
0.0263
1


Household
0.0875
0.0720
0.0141
0.0360
- 0.0960
- 0.0694
1


Note: Livloc = living location, Foodacc = Access to food, Elecuse = Electronic use, Empstatus = Employment status


Appendix C
Regression Statistics
 
 
 
 
 

Multiple R
0.33652961




 

R Square
0.113252178




 

Adjusted R Square
0.065059362




 

Standard Error
16.8938651




 

Observations
98




 

 





 

ANOVA





 

 
df
SS
MS
F
Significance F
 

Regression
5
3353.454
670.6907
2.34998
0.04690976
 

Residual
92
26257.05
285.4027


 

Total
97
29610.5
 
 
 
 

 





 

 
Coefficients
Standard Error
t Stat
P-value
Lower 95%
Upper 95%

Intercept
75.59809887
5.003943098
15.10771
1.18E-26
65.65983595
85.53636179

Livloc
2.253329357
3.634280783
0.620021
0.536777
-4.964665978
9.471324693

Foodacc
-7.64257711
3.467251423
-2.2042
0.030005
-14.52878832
-0.759267103

Age
-0.571668292
0.4892296111
-1.16835
0.245685
-1.543452603
0.40011602

Gender
-4.267752145
3.482703453
-1.22541
0.223547
-11.18470182
2.64919753

Elecuse
7.374578497
3.508810443
2.101732
0.03831
0.405778087
14.34337891


Table C1: Regression Results of Equation 1



Appendix D
Regression Statistics
 
 
 
 
 

Multiple R
0.336793902




 

R Square
0.113430133




 

Adjusted R Square
0.054974976




 

Standard Error
16.9847304




 

Observations
98




 

 





 

ANOVA





 

 
df
SS
MS
F
Significance F
 

Regression
6
3358.722939
559.7871566
1.940464111
0.082602457
 

Residual
91
26251.77706
288.4810666


 

Total
97
29610.5
 
 
 
 

 





 

 
Coefficients
Standard Error
t Stat
P-value
Lower 95%
Upper 95%

Intercept
75.36608084
5.315703569
14.1780067
9.13302E-25
64.80708871
85.92507297

Livloc
2.224791159
3.65992455
0.6078789968
0.544781856
- 5.045199354
9.494781673

Foodacc
- 7.674850869
3.4940951118
- 2.196520303
0.030595759
- 14.61544159
- 0.734260151

Age
- 0.57180219
0.491928836
- 1.162367702
0.248125499
- 1.548958391
0.40535401

Gender
- 4.282903174
3.503229668
- 1.22255849
0.224653485
- 11.24163855
2.675832204

Elecuse
7.414099832
3.539782288
2.094507297
0.038994363
0.382757164
14.4454425

Household
0.469898688
3.476848599
0.135150748
0.892791184
-6.436433937
7.376231314


Table D1: Regression Results for Equation 2


Appendix E
Regression Statistics
 
 
 
 
 

Multiple R
0.342072147




 

R Square
0.117013354




 

Adjusted R Square
0.058794454




 

Standard Error
16.95037232




 

Observations
98




 

 





 

ANOVA





 

 
df
SS
MS
F
Significance F
 

Regression
6
3464.823919
577.4706531
2.009886043
0.072303175
 

Residual
91
26145.67608
287.3151218


 

Total
97
29610.5
 
 
 
 

 





 

 
Coefficients
Standard Error
t Stat
P-value
Lower 95%
Upper 95%

Intercept
76.93724511
5.462019143
14.08586149
1.37562E-24
66.08761507
87.78687515

Livloc
2.386240694
3.65268057
0.6532848
0.515219931
- 4.869360543
9.641841932

Foodacc
- 7.629873326
3.478908187
- 2.193180422
0.030843471
- 14.54029707
- 0.719449581

Age
- 0.600289213
0.493080337
- 1.217426792
0.22658947
- 1.57973273
0.379154304

Gender
- 3.832991673
3.563443313
- 1.075642668
0.284930698
- 10.91133406
3.245350715

Elecuse
7.449355935
3.522595001
2.114735283
0.037186188
0.452153701
14.44655817

Empstatus
- 2.307469864
3.706214875
- 0.622594734
0.53510793
- 9.669410422
5.054470693


Table E1: Regression Results for Equation 3


Appendix F
Regression Statistics
 
 
 
 
 

Multiple R
0.342175813




 

R Square
0.117084287




 

Adjusted R Square
0.048413065




 

Standard Error
17.0435963




 

Observations
98




 

 





 

ANOVA





 

 
df
SS
MS
F
Significance F
 

Regression
7
3466.924
495.2749
1.704998
0.11776
 

Residual
90
26143.58
290.4842


 

Total
97
29610.5
 
 
 
 

 





 

 
Coefficients
Standard Error
t Stat
P-value
Lower 95%
Upper 95%

Intercept
76.77540486
5.812501355
13.20866873
8.82E-23
65.2278564
88.3295332

Livloc
2.366687686
3.679960974
0.643128474
0.521776
- 4.944197092
9.677572463

Foodacc
- 7.650487753
3.506432233
- 2.181843892
0.031726
- 14.6166274
- 0.684348106

Age
- 0.600056056
0.495799772
- 1.210279009
0.229341
- 1.58504884
0.384936727

Gender
- 3.847418552
3.587056276
- 1.072583828
0.286325
- 10.97373193
3.278894828

Elecuse
7.4735584226
3.553386272
2.103221506
0.038237
0.414136387
14.53298046

Empstatus
- 2.281834597
3.73877298
-0.610316435
0.54319
- 9.70955969
5.145890495

Household
0.297638949
3.500296748
0.08503249
0.932424
- 6.656311486
7.251589383


Table F1: Regression Results for Equation 4