Project analysis -22 dec 2014-am (1)

Haruna Emmanuel P73270

Suresh Mani P77104

Ashok Sivaji P77800

Dwi Budiningsari P75375

Hamzah Wali P74918

Nadzirah Hanis Zainordin P75182

Marwan Jalambo P75376

Ooi Theng Choon P75129

Presenters

NNPD6014 – Group 1

Introduction Research Question Objective Research Hypothesis Methodology

◦ Research Design◦ Study Population◦ Study Population and Sampling Frame◦ Sampling Size◦ Sampling Methodology◦ Inclusion and Exclusion◦ Tools & Instrument◦ Reliability and Validity

Analysis Two-way ANOVA between group

Binary Logistic Regression

Conclusion References

Nutrition is an important factor in the performance and

health of human.

80% of adults do not know recommended calorie level(US

National surveys in US (2011)

20% of students are aware of recommended calorie and

40% overweight (Yu-Chieh et al 2012).

Diseases vary by races and gender (Rajakumar , 2012) )

In addition to the population breakdown,

another reason for this is because

1. Malays reported highest percentage of

hypertension 17% followed by Chinese 11%

(Rajakumar , 2012)

2. Chinese reported lowest health awareness

(Rajakumar , 2012)

Adults, 18 years and above

◦ 33.3% (5.4 million) are pre-obese

◦ 27.2% (4.4 million) are obese

Children below 18 years (based on weight for

age status)

◦ 3.9% (0.3 million) are obese

Source: NATIONAL HEALTH AND MORBIDITY SURVEY 2011

Transition from age related disease to Lifestyle related disease

Fifteen percent of Malaysians above the age of 18 years are diabetics.

The prevalence among those above 40 years old was 17.6%,

Hypertension: 31.6% once they reach age 55 years old

Hypertension 35.8% in above 40 years

Source: NATIONAL HEALTH AND MORBIDITY SURVEY 2011

Can gender, age, race, BMI, education level and family history predict knowledge scores on calorie and BMI among students and staff of Faculty of Health Sciences, UKM ?

To determine interaction between race and school

on the knowledge scores on calorie and BMI among

FSH student and staff

To determine the influence of gender, age ,race,

BMI, education level and family history in predicting

knowledge scores on calorie and BMI among FSH

student and staff

Gender, age, race, BMI, education level and family history influence knowledge scores on calorie and BMI among students and staff of Faculty of Health Sciences, UKM.

Race

Age

BMI

Gender

Education

Family History

Knowledge Scores on Calorie and BMI

Cross sectional study using a web-based survey questionnaire

Period:◦ October – December 2014

Location:◦ Faculty of Health Science, UKM, Jalan Raja Muda

Abdul Aziz, Kuala Lumpur, Malaysia

Study Population◦ Students and staff of FHS

Sampling Frame◦ List of students and staff currently enrolled and

employed in FHS respectively

For a known population of N=1500, ( Krejcie R V & Morgan D W. , 1970)

For a known population of N=1500, ( Krejcie R V & Morgan D W. , 1970)

n = 3.841 (1500) (0.5) (1-0.5)0.052 (1500-1) + 3.841 (0.5) (1-0.5)

= 1440.375

3.75 + 0.96025

= 1440.375

3.81

n = 378

Random Sampling

Target of minimum 31 samples (n > 30) from each schools◦ Diagnostic & Applied Health Sciences (DAHS)◦ Healthcare Sciences (HS)◦ Rehabilitation Sciences (RS)

Corder, G. W., & Foreman, D. I. (2009, p.2) The minimum sample size for using a parametric statistical test varies among texts. For example, Pett (1997) and Salkind (2004) noted that most researchers suggest n>30. Warner (2008) encouraged considering n>20 as a minimum and n> 10 per group as an absolute minimum."

Survey was sent to 150 samples randomly from each schools◦ 278 responses received out of 450 (62 %)◦ Complete questions response was obtained from

179 /450 ( 40%)

n = 179

Comparison with calculated sample size◦ 179/378, (47%)

33%

45%

22%

Breakdown of Respondents by Schools

within FHS

Diagnostics &

Applied Health

ScienceHealthcare

Science

Rehabilitation

Science

33% => 59 responses

45% => 81 responses

22% => 39 responses

Inclusion ◦ Both full & part time students enrolled

◦ Staff employed by FHS

◦ Malaysian citizen

◦ Responses received from 25th Nov 2014

Exclusion◦ Incomplete survey items

◦ Student / staff on study leave

◦ Response received after 5th Dec 2014, 6.30pm

Tool◦ Use Mi-UXLab version 1.0◦ http://usability.mimos.my/miuranus/survey/login.

php?key=rE6xWTOx34pfn8k82Fk4FT41tOwoh5◦ 38 Items

Part A Knowledge on Calorie ( 23 questions)

Part B BMI (5 questions)

Part C Demographic (10 questions)

◦ Correct answer is given one mark, wrong answer is given no marks

◦ Part A & B : Each question has 5 options (MCQ)◦ Score > 70% rated as High Score

Mean = 53%IQR = 17High Score = Mean + IQR = 70%

Reliability◦ Reliability was done using Cronbach/Coefficient alpha

◦ For both Test 1 and Retest 2, Cronbach Alpha,

◦ α > 0.7 = > Highly reliable questionnaire items

◦ Paired-Sample t-test t(23) = -0.591, p=0.560

◦ Since p > 0.05, there is no significant difference between Test 1 and Retest2

Cronbach Alpha N of items

Test 1 0.771 28

Retest 2 0.714 28

Validity◦ Face validity

Readability Statistics Score / Grade

Flesch Reading Ease 55.0FairlyDifficult

65.3 (Standard)

Flesch Grade Level 8.7 6.7

Validity◦ Face validity

Readability Statistics Score / Grade

Flesch Reading Ease 55.0FairlyDifficult

65.3(Standard)

Flesch Grade Level 8.7 6.7

Expected due to technical and localized terms used in survey such as ‘ikan kembong’, saturated, unsaturated, monosaturated, polysaturated, trans-fatty acid, ‘otak-otak’, calorie, BMI

Two Way ANOVA Between Groups

Binary Logistics Regression

Use SPSS version 22.0

School Mean SE

Malay 50.73% 1.40%

Non-Malay 55.21% 1.45%

School Mean SE

Diagnostic 47.67% 1.39%

Healthcare 60.13% 1.48%

Rehabilitation 44.41% 1.76%

• There is no interaction effect for School * Race , F ( 2,173)=0.093 (p=0.911)

• We look at main effect, both school and race are significant p < 0.05, post hoc analysis required

According to Cohen (1998), for both school and race, effect size is small r < 0.29 ,

• The mean score from Diagnostic and Healthcare schools significantly differ ( p < 0.05)

• The mean score from Rehabilitation and Healthcare schools significantly differ ( p < 0.05)

• Sample size for schools are widely different, use post-test Hochberg’s GT2

In order to determine the interaction between race and school on the knowledge

scores on calorie and BMI among FSH student and staff, a Two way between groupsANOVA was performed.

Prior to interpreting the results of Two-way between group ANOVA, severalassumptions were evaluated. First, the Cook’s distance obtained from the residualstatistics ( < 1) shows that each independent variable is normally distributed. Second,the Levene’s test is not significant (p > 0.05) implying the population variances forschool and race is approximately equal, hence homogeneity of variance is assumed.Descriptive statistics , means of independent variables are shown in the tables.

Dependent Variable Independent Variable Mean Std Dev.

% Score by

Respondents

Diagnostics School 47.68% 1.39%

Healthcare School 60.13% 1.47%

Rehabilitation School 44.41% 1.76%

Dependent

Variable

Independe

nt Variable

Mean Std Dev.

% Score by

Respondents

Non-Malay 55.21% 1.45%

Malay 50.73% 1.40%

The test of between subject effects, shows that there is no interaction effect forSchool * Race , F ( 2, 173)=0.093 (p=0.911). However, for the main effects, bothschool and race are significant p < 0.05, post hoc analysis required. Since the samplesize for schools are widely different, Hochberg’s GT2 posttest was used. From this, it wasfound that the mean score from Diagnostic and Healthcare schools significantly differ;and the mean score from Rehabilitation and Healthcare schools significantly differ.

-2 Log Likelihood Cox & Snell R Square Nagelkerke R Square

88.513 0.167 0.339

Chi-square df Sig.

6.101 8 0.636

Odds Ratio (estimated)

Lower Upper BLR (Predictor Ratio)

Race(NonMalay / Malay)

1.091 0.416 2.861 1.490

Gender(Female/Male)

1.231 0.337 4.496 1.664

Education(Graduate / Undergraduate)

18.167 4.044 81.615 15.310

Family Disease (Family History / No Family History)

1.966 0.675 5.723 1.412

Two – way ANOVA

Binary Logistic Regression

There are no interaction between race and school

on knowledge score on calorie and BMI among

FSH student and staff. However main effects,

school and race have significant effect on

knowledge score

Education made a significant contribution in

predicting knowledge scores on calorie and BMI

among FSH student and staff

There is no interaction effect (p<.05) between School * Race F ( 2,173)=0.093 (p>.05)

We can see from the Test of Between-Subjects Effects table that ◦ for our factor “School”, we have found a

significant effect, F (2, 173) = 30.698, p < 0.05.◦ for our factor “Race”, whether Malay or non-

Malay we have found a significant effect, F (1,173) =7.926, p < 0.05

A logistic regression analysis was conducted to predict the knowledge level on calorie and BMI. This was coded as (High =0 and Low = 1) while the independent variables were Education, Gender, Race, Family history of disease, BMI and Age.

A test of the full model against a constant only model was statistically significant, indicating how well that the predictors influence the prediction. (chi square = 32.627, p < .001 with df = 6).

Nagelkerke’s R2 of .339 indicated a moderately weak relationship between prediction and grouping. Prediction success overall was 88.8%.

The Wald criterion demonstrated that only education made a significant contribution to prediction (p = .002). Other variable were not significant predictors.

EXP(B) value indicates that when education level is raised by one unit the odds ratio is 15 times as large and therefore graduate are 15 more times likely to be more knowledgeable on calorie and BMI than undergraduate.

1. Sivaji, A., & Tzuaan, S. S. (2012, July). Website user experience (UX) testing tool development using Open Source Software (OSS). In Network of Ergonomics Societies Conference (SEANES), 2012 Southeast Asian (pp. 1-6). IEEE.

2. Chuan, N. K., Sivaji, A., Shahimin, M. M., & Saad, N. (2013). Kansei Engineering for e-commerce Sunglasses Selection in Malaysia. Procedia-Social and Behavioral Sciences, 97, 707-714.

3. Sivaji, A., Shopian, S., Nor, Z. M., Chuan, N. K., & Bahri, S. (2013). Lighting does Matter: Preliminary Assessment on Office Workers. Procedia-Social and Behavioral Sciences, 97, 638-647.

4. Elliott, L. J. (2013, September). Teaching Undergraduates about Human Factors through Research in Usability. In Proceedings of the Human Factors and Ergonomics Society Annual Meeting (Vol. 57, No. 1, pp. 487-490). SAGE Publications.

5. “Sample Size Calculator, ”http://www.surveysystem.com/sscalc.htm

6. “Schools within FSK” http://www.fsk.ukm.my/?page_id=890#

7. https://statistics.laerd.com/spss-tutorials/two-way-anova-using-spss-statistics.php

8. Survey Preparation Checklist, http://academy.cultureamp.com/survey-prep-checklist/

9. SPSS version 20.0 for Windows, Analysis without Anguish, Sheridan J Coakes

10. Andy Field, “Discovering Statistics using IBM SPSS Statistics 4th Edition”

11. Krejcie, R. V., & Morgan, D. W. (1970). Determining sample size for research activities. Educ Psychol Meas.

12. Corder, G. W., & Foreman, D. I. (2009, p.2). Nonparametric statistics for non-statisticians: a step-by-step

approach. New Jersey: Wiley & Sons.

13. International Food Information Council Foundation (IFIC), “2011 Food and Health Survey,” 2011

14. Yu-Chieh, C., Debra B, R., Natalia, V., & Shu, W. (2012). University Student Sample Is Unable to Accurately

Assess Their Calorie Needs: Implications for Weight Management and Menu Labeling. Food and Nutrition

Sciences, 2012.

15. A Aylin Alsaffar “Validation of a general nutrition knowledge questionnaire in a Turkish student sample” Public

Health Nutrition, 15(11), 2074–2085, 2012.

16. Aida Malek Mahdavi, Paria Abdolahi, Reza Mahdavi “Knowledge, Attitude and Practice between Medical and Non-

Medical Sciences Students about Food Labeling” Health Promotion Perspectives, 2(2); PP: 173-179, 2012.

17. Cupisti, C. D’Alessandro, S. Castrogiovanni, A. Barale, and E. Morelli, “Nutrition knowledge and dietary

composition in Italian adolescent female athletes and non-athletes”, International Journal of Sport Nutrition,

12(2)207–219,2002.

18. David W Haslam, W Philip T James “Obesity” The Lancet 366 (9492) pp: 1197-1209, 2005.

19. Geok-Lin Khor “Food availability and the rising obesity prevalence in Malaysia” IeJSME, 6 (1): 61-68, 2012.

20. Hamid Arazi, Rastegar Hosseini “A Comparison of Nutritional Knowledge and Food Habits of Collegiate and Non-

Collegiate Athletes” SportLogia, 8(2) pp: 100-107, 2012.

21. Jessie A. Satia, Joseph A. Galanko, Marian L. Neuhouser “Food nutrition label use is associated with demographic, behavioral, and psychosocial factors and dietary intake among African Americans in North Carolina” Journal of the American Dietetic Association, 105 (3), PP 392–402, 2005.

22. Lauren B. Headrick, Cassie C. Rowe, Ashley R. Kendall, Michelle A. Zitt, Dawn L. Bolton, Bobbi Langkamp-Henken, “Adults in All Body Mass Index Categories Underestimate Daily Energy Requirements” Journal of Nutrition Education and Behavior, 45 (5), 2013

23. M. Azizi, F. Rahmani-Nia, M. Malaee, M. Malaee, and N. A. Khosravi, “A study of nutritional knowledge and attitudes of elite college athletes in Iran” Brazilian Journal of Biomotricity, 4,105–112, 2010.

24. Marquitta C. Webb and Safiya E. Beckford, “Nutritional Knowledge and Attitudes of Adolescent Swimmers in Trinidad and Tobago”, Journal of Nutrition and Metabolism, 2014, Article ID 506434, 7, 2014.

25. Wan Nazaimoon Wan Mohamud, Kamarul Imran Musa, Amir Sharifuddin Md Khir, Aziz al-Safi Ismail, IkramShah Ismail, Khalid Abdul Kadir, Nor Azmi Kamaruddin, Nor Azwany Yaacob, Norlaila Mustafa, Osman Ali, Siti Harnida Md Isa, Wan Mohamad Wan Bebakar “Prevalence of overweight and obesity among adult Malaysians: an update” Asia Pac J Clin Nutr, 20 (1):35-41, 2011.

26. Joffe, S., Cook, E. F., Cleary, P. D., Clark, J. W., & Weeks, J. C. (2001). Quality of informed consent in cancer clinical trials: a cross-sectional survey.The Lancet, 358(9295), 1772-1777.

27. Flesch R: A new readability yardstick. J Appl Psychol 1948, 221-233.

28. Rajakumar, S., Ann, L. L., Gill, P. K., Xin, G. W., & Kalasalingam, A. (2012) A comparative study of hypertension, diabetes mellitus and obesity among Malaysians in urban regions-A cross sectional study. Malay, 664, 72-6.