Relationships between parents’ academic backgrounds and incomes and building students’ healthy eating habits

Research objectives and hypotheses

To achieve the aim, the following objectives and hypotheses have been formulated:

1. To understand primary school students’ knowledge, attitudes and practices on building healthy eating habits.

2. To find the difference of knowledge, attitude and practices regarding eating habits in terms of students’ gender.

3. To find difference of knowledge, attitude and practices of building healthy eating habits of students in terms of parents’ academic background.

4. To find the difference of knowledge, attitude and practices of students’ healthy eating habits in terms of parents’ income.

Research Hypotheses

• Ho1: There is no difference of knowledge, attitude and practices regarding healthy eating habit in terms of students’ gender;

• Ho2: There is no difference of knowledge, attitude and practices of students’ healthy eating habit in terms of parents’ academic background; and

• Ho3: There is no difference of knowledge, attitude and practices of students’ healthy eating habit in terms of parents’ income.

Research design

The research design was developed to address the research objectives and hypotheses of the study which involved the quantitative method. The current research is a cross-sectional study performed among primary schools in Kulim, Kedah Darul Aman in April 2016. Generally, the research method of this study is a modification of the methods used in the report by Central Health Education Unit (2009), Department of Health of Hong Kong. The questionnaires are adapted from the report for each category of respondents and modified to suit the local standard.

Population and sample

The population of this study consists of all the primary schools students in Kulim district, Malaysia for the academic year 2016. Kulim district is chosen due to urban and sub-urban area as there is another study by the corresponding author (Hoque et al., 2016) in the capital city. As the previous study was conducted in the capital city, this study was intentionally conducted through convenient sampling method in an urban and semi-urban area. There are differences among city and urban students in terms of food consumption and habit (Noraziah & Mohd Azlan, 2012). This study is also included students’ gender, parents’ education and their income as new variables. There are about 50 schools in this district of which only 18 have intended to participate. The sampling of schools is based on their willingness to participate the study. A request letter is sent to all 50 primary schools in the whole of Kulim district (338 km north–west from the capital Kuala Lumpur, Malaysia) for which the list of schools are obtained from the Ministry of Education website. Therefore, the respondents are chosen from 18 schools unlike the study by Central Health Education Unit (2009) in which schools were chosen based on stratified cluster sampling method. The other schools disagree due to their students’ examination preparation and other businesses. But the principals have not clarified what other businesses they have had. In each selected school, a number of students are chosen randomly. Each respondent is given a set of survey questions to be filled depending on their categories.

Selection of students

For the ease of answering survey questions, only students’ age 10 and 11 are chosen to answer questions regarding their dietary habits. Students younger than 10 years old are not preferred since the questions might be difficult for them. Students aged 12 years old are not selected to allow them to prepare for their upcoming Primary School Achievement Test (Ujian Pencapaian Sekolah Rendah). The total number of students for this study is set at 360 students taking 20 students for each school. Altogether 10 students are chosen randomly from each class (year 4 and year 5). Each student of the class is given an identification number. All of these same identification numbers (written each in a small piece of paper) are also put in a box. The class teacher chooses 10 numbers one by one from the box. Thus 10 students from each class are chosen by lottery. The questionnaires are distributed to the students and recollected by the class teacher after seven days and then are passed back to the researcher for further analysis. Finally, data from 318 students are analysed due to either missing information or received after due date. The missing information is mostly about parents’ education level and income due to their parents’ unwillingness to disclose their personal information. Sample size is calculated using the formula (1/√N, where N is the number of participants or sample size) to justify the power which shows the significance level at 0.05 up to two decimal places (1//√318 = 0.05). It means, there is only 5% chance of differing the result for sampling error from the actual population (Niles, 2006).

Research instruments

In this study, a survey questionnaire is administered by the researcher. The questionnaire is used to gather the quantitative data from the participants. The researcher has considered the survey questions to be sufficient to provide data regarding dietary habits. The questionnaires are designed to gather only answers that can easily be quantified and analysed through data analysis procedures. All of the questions are written in Malay language.

Ethical considerations

The questionnaires are assured to contain no prejudice, racial stigmatization, demeaning nature, cynical notion and sarcasm or anything provocative, questions that instigate, or implying of incompetent parenting, mismanaged health and nutrition of students and so forth. The questions are thoroughly screened and rephrased to avoid misunderstanding of the intention of the questions. The questions are also written in a very objective manner, simple and straightforward, clear and specific. Aside from that, the researcher also respects the school’s privacy and decisions during the survey. It is pertinent that the researcher comply with the rules given by the school to ensure smooth response from the participants. The permission is taken from University of Malaya Research ethic Committee before conducting the survey (UM.TNC2/RC/H&E/UMREC-45).

Data collection procedure

The researchers have performed the survey by approaching individually the schools on Sundays (working day in Kulim, different than other part of Malaysia) during schooling hours, starting from 8 a.m. until 3 p.m. basically; the work flow of the researcher is simplified in the following arrangements:

1. Enter school (walk-in).

2. Request permission with an official letter.

3. Pass the survey questions for students and parents to the administrative staff.

4. Leave the school.

5. Collect survey questions after a week.

The procedure in collecting data is simple and direct and does not require waiting for approval. Schools that refused to cooperate are not included in the study. An immediate approval from supportive schools is highly appreciated and expedited the data collection process.

Data analysis

There are two main types of statistics used in this study: descriptive and inferential statistics.

Demographic information of students

The demographic information of the students is presented in Table 1, which shows the year group, gender, ethnicity, parents’ academic background and income. The data shows that the Year 5 students are found to be the highest number of respondents in this research, which is 186 of them (58.5%) as compared to the Year 4 students who are only 132 (41.5%). As usual, the highest respondents are the female student which is 171 of them (53.8%) compared to the male students who are only 147 (46.2%). The data also shows that the respondents are from three ethnicities. The highest ethnicity which the respondents gathered is Malay with 230 (72.3%) and followed by Indian with 67 respondents (21.1%). The Chinese respondents are the least with only 21 of them (6.6%). The parents’ academic background item is categorized into four categories which are primary school, secondary school, matriculation and college or higher learning institution. The data shows that highest is secondary school with 156 respondents (49.1%), followed by college or higher learning institution, 120 respondents (37.7%) and primary school with 34 respondents (10.7%). The least is parents with matriculation academic background, eight of them (2.5%). The highest respondents for family income less than RM3999 are 182 (57.2%), followed by RM4000-RM7999 with 100 (31.4%) and range from RM8000-RM9999 has 21 respondents (6.6%). The least respondents (15) are from family income of RM10000 and above (4.7%).

Knowledge, attitudes and practices of school students on healthy eating habit

Knowledge, attitudes and the practices of primary school students on healthy eating is presented below.

Knowledge of school students on healthy eating habit

In identifying knowledge about healthy food, six pairs of food are given in the questionnaire. Students are asked to choose the healthiest food among the six pairs. Based on the Table 2, it shows that the student are able to identify the healthy food choices. The food choice which has the highest rate are the correct choice of healthy foods like soya sauce drumstick with 270 respondents (84.9%), orange juice with 315 respondents (99.1%), raisin bread with 304 respondents (95.6%), yoghurt with 286 respondents (89.9%) and spaghetti with fresh tomato sauce with 280 respondents (88.1%). The last choice has a different result as 280 respondents (88.1%) who have selected fried noodles with chicken as healthy food instead the fried noodles with vegetables which has only 38 respondents (11.9%). This shows that the students are a little confused and believe that chicken is much healthier compared to vegetables.

Students’ food preference attitude

In identifying students’ preference in food, the same six pairs of food and drinks are given in the questionnaire. Students are asked to choose the food choices which they prefer among the six pairs. Based on the Table 3, it shows there is a difference in answers between students’ preference and their perceived opinions on what healthy food should be. The percentages in students’ food preferences are generally lower. For instance, although many students perceive deep fried drumstick, ice-cream and carbonate drink as unhealthy foods, more than half of the respondents still prefer eating these foods despite the fact that they are an unhealthy choice. Food like deep fried drumstick has higher score of 161 respondents (50.6%) while soya sauce drumstick has only 157 respondents (49.4%) and carbonated drink has the higher respondents of 255 (80.2%) while orange juice has only 63 respondents (19.8%). Raisin bread has 210 respondents (66%) which is higher than hotdog which has 108 respondents (34%), yoghurt with 91 respondents (28.6%) while ice cream has higher score of 227 respondents (71.4%) and spaghetti with fresh tomato sauce has 169 respondents (53.1%) while the hamburger and fries scored less with only 149 respondents (46.9%).

Practices—students’ daily dietary habit

Students’ are required to identify their daily dietary habit in the week prior to the questionnaire being carried out.

According to Table 4, the percentage of students consuming fruits, vegetables and dairy products more than twice are on an average level between 54.4% (173 respondents), 46.2% (147 respondents) and 52.8% (168 respondents). Grains are also consumed more than twice with 68.9% (219 respondents). However, there is a high average of consumption in unhealthy eating such as having deep fried food 34.9% (111 respondents) and drinks with added sugar with 50.6% (161 respondents) once in a week. Foods that are high in sugar, salt and fat are consumed once in a week. Those who do not reach the recommended frequency stand between 52.8% (168 respondents), 41.8% (133 respondents) and 43.7% (139 respondents). From this result, it is understood that the students are having a positive attitude towards their daily dietary habits.

Differences of knowledge, attitude and practices based on students’ gender

H_O1: There is no significant difference of knowledge, attitude and practices based on students’ gender.

An independent samples t-test (Table 5) is conducted to compare the mean of knowledge, attitude and practices for males and females student. Based on the p value for knowledge (p = 0.76), attitude (p = 0.68) and practices (p = 0.50), this study finds that the assumption of p < 0.05 is not met. Thus, the null hypothesis H_O1 is accepted. This finding indicates that there are no significant differences of knowledge, attitude and practices based on student’s gender.

Differences of knowledge, attitude and practices based on parents’ academic backgrounds

H_O2: There is no significant difference of students’ knowledge, attitude and practices of healthy eating habit based on parents’ academic backgrounds.

A one-way ANOVA between groups analysis of variance (Table 6) is conducted to explore the mean differences of knowledge, attitude and practices based on four group of parent’s academic background. In Table 6, the F distribution for knowledge (2.49), attitude (0.42) and practices (1.22) are smaller than the critical F value 2.42. Therefore, the H_O2 cannot be rejected. Based on the p value for knowledge (p = 0.60), attitude (p = 0.74) and practices (p = 0.30), this study finds that the assumption of p < 0.05 is not met. Thus, the null hypothesis H_O2 is accepted. This finding indicates that there are no significant differences of knowledge, attitude and practices of students’ eating habit based on four group of parent’s academic background.

Differences of knowledge, attitude and practices based on parents’ incomes

H_O3: There is no significant difference of knowledge, attitude and practices of students’ eating habit based on parents’ incomes.

A one-way ANOVA between groups analysis of variance (Table 7) is conducted to explore the mean differences of knowledge, attitude and practices based on four group of parent’s income. Based on the p value for knowledge (p = 0.015), attitude (p = 0.44) and practices (p = 0.58), this study finds that only the comparison of knowledge is less than 0.05 (F = 3.57 which is higher than critical value 2.42, df (3, 314)); while for the other two constructs F value are 0.91 for attitude and 0.66 for practices which are smaller than the critical F value of 2.42. Accordingly, the assumption of p < 0.05 is not met. It means that the null hypothesis H_O3 is partially rejected. Result shows, there is a significant difference in mean knowledge for the four income group while there are no differences in mean attitude and practices of building students healthy eating habits in terms of parents’ income differences. The post hoc LSD (Table 8) shows that the significance difference occurs between 8,000 and 9,999 income group and the other three groups.

Students’ knowledge of healthier food choice

The study finds most students having a good knowledge about healthy food choices. For example, in the selection of healthy drinks, 315 students (99.1%) have managed to pick orange juice as healthier than carbonated drinks and yoghurt as well as the selection of healthier beverages than ice cream (89.1%). A similar situation identified in the selection of healthy food, where 270 students (84.9%) have chosen the soy sauce drumstick instead of choosing deep fried drumstick. This means that students understand well the definition of healthy food and the types of food that are considered healthy. This finding is aligned to Lin et al. (2008) in Taiwan and Hoque et al. (2016) in Malaysia who have also found that primary school children acquire simple knowledge in nutrition. However, the percentage of students who has answered the test correctly is only at the rate of 67–71%. Among the other factors that influence students’ knowledge about nutrition and healthy eating is the influence of family and the media. Dixey et al. (2001) explains that most children already understand about balanced diet and its relation to health. In addition, the influence of peers has also helped them understand about healthy eating. This shows that most students are generally aware of the necessity of different food groups, the principle of balanced diet, the skill of selecting foods to achieve the balanced, and the comparison of foods in terms of specific nutrients. Children in general value the importance of nutrition, but they are not concern of the health benefit of foods in food selections (Lin et al., 2008). The study reports that students have good knowledge in the pre-test, before the intervention starts. This situation indicates that school children have already existing knowledge about healthy eating. This is aligned with Piperakis et al. (2004) finding which reveals students’ understanding on the relationships between healthy eating and physical and mental development.

Knowledge, attitude and practices based on students’ gender

The study finds no significant difference of knowledge on students’ healthy eating habit based on students’ gender. This finding contradicts with Naeeni et al. (2014) and Skardal et al. (2014) who have found that there is a tendency for girls to have a healthier diet than boys, with greater intake of fruits and vegetables (girls intake in median 3.5 units per day and boys 2.9 units per day), and lower intake of soft drinks (girls 0.25 L in median versus boys 0.5 L per week). But Lin et al. (2008) have found no statistically significant differences between genders for nutrition knowledge.

The study also finds no significant difference of attitude and practices of students’ healthy eating habit based on students’ gender. This result is again congruent to Lin et al. (2008), with no statistically significant differences between genders for nutrition attitude and practices. But Petrauskienė, Žaltauskė & Albavičiūtė (2015) have contrast result of significant differences in terms of attitude and practices in breakfast consumption; girls eat breakfast less frequently than boys.

Even the study has found no significant differences by gender, yet a study shows that there are no gender differences in number of foods tried, liked or disliked, although the significant age-by-gender interaction in number of foods disliked suggests that boys may be more ‘picky’ early, where girls are more so during adolescence (Cooke & Wardle, 2005). Another study reports that girls appear to have healthier eating patterns than boys. This is most evident in their consumption of fruit and vegetables (Bradshaw & Benton, 2012).

Knowledge, attitude and practices based on parents’ academic backgrounds

The study reports no significant difference between knowledge, attitude and practices of building healthy eating habit in terms of parents’ academic backgrounds. This fact is verified in other research (Manios et al., 1999). The child’s broader social environment, including that of the school (Kennedy, Meyers & Layden, 1996) and the family (Mandell, 1993; Brown & Ogden, 2004) is recognized as an important influence in supporting children’s healthy eating behaviour.

There is found no significant difference based on parents’ academic backgrounds. However, children of mothers with a high educational level consume more pieces of fruit per day, more grams of vegetables per day and are more likely to have breakfast on a daily basis than children of mothers with a low educational level. These findings are in line with van Ansem et al. (2014) who have figured out that parental breakfast consumption mediated the association between maternal education level and children’s breakfast consumption. A contradict result shows that parental educational level and parental unemployment are unrelated to adolescents and children’s breakfast consumption (Pearson, Biddle & Gorely, 2009). A recent study (Vollmer & Baietto, 2017) has found negative association of parental control and children food consumption.

Knowledge, attitude and practices based on parents’ incomes

The study has revealed a significant difference of knowledge of students’ healthy eating habit based on parents’ incomes. This result is slightly congruent to Venezuelan study (Moya De Sofontes & Dehollain, 1986) of affecting children knowledge by family’s shopping pattern and a tight family budget (Ling, Robbins & Hines-Martin, 2016). Another study shows that parents with higher socio-economic status have reported a significantly better knowledge of dietary guidelines compared to those with lower socio-economic status (Skardal et al., 2014). Though, in this study, no significant difference was found in students’ attitude toward healthy eating habit in relation to parents’ incomes, Nelson (2000) has a totally different finding which explains that healthy foods are frequently more expensive, families in poverty live options or families on lower incomes differ in terms of their education and food culture, leading them to make less healthy food choices in areas with lower availability of healthy food. The result related to students healthy food practices of this study also contradict to some other studies (Skardal et al., 2014; Roos et al., 1999) as they have reported that students from families with higher socio-economic status have reported a significant higher intake of vegetables and fish, and lower intake of soft drinks and fast food than those from lower social economic status. It is evident that there is a strong link between the brands children eat and what their family can afford (Ludvigsen & Sharma, 2004) and has a direct association with children’s academic achievement, health status and parents’ income (Abdollahi et al., 2008; Opoola, Adebisi & Ibegbu, 2016). Evidence shows that parents have expressed a desire for nutrition classes including whole family in which topic areas of interest included guidelines of purchasing and cooking healthier foods, knowing the ways to encourage their children to eat healthier foods and knowing the ways of food labels (Prelip et al., 2011).

As a conclusion, parents’ income levels significantly influence students’ knowledge of building healthy eating habits as parents usually buy the foods based on their monetary capability rather to healthy food knowledge. If parents keep healthy food available at home and model healthy eating in front of them then it influences children’s eating habit (Vollmer & Baietto, 2017).

Limitations and delimitations

Limitations of the study include the respondents, time, location of survey and self-reported questionnaire. Since the study is performed in Kulim district, the data obtained may not be representative of the students in Kedah state (in which the Kulim district is situated) much less to whole Malaysia. Kulim district is also known to be an area populated mostly by medium income families. Data obtained would generally be influenced by the income status and may show similar tendencies. The students who have participated in the survey are only 10 and 11 years old. They are instructed to take the questionnaire at home and collect parents’ income and education information. It is not sure whether they have taken any other information out of instruction. Furthermore, the survey may produce some obstacles in obtaining data at the same time since more than 10 schools are involved. The study is delimited by the questions that are possible for questioning by comparing to the status of the respondents. Deeper excavation of other pertinent information that influences dietary pattern may not be possible through a survey such as food tolerance, clinical issues, and family tradition and so forth. A mixed method study could be more effective for in depth understanding of the problem.