Validity and Reliability of Knowledge and Behavioral Questionnaire about Weight Loss Diet in Teenage Girls

Background: The choice of diet method by teenage girl can be influenced by the knowledge. Methods: This study aimsto determining the validity and reliability of the knowledge and diet behavior questionnaire. The subject of this research is 30 female students. Data analysis of the validity using analysis of difficulty levels, differentiating power, distractor analysis, discrimination test and Lawshe’s Content Validity Ratio (CVR) analysis. The reliability test using Cronbach’s Alpha Coefficient test. Results: The results showed that 24 items of knowledge questions were declaired valid and 15 items in the diet behavior questionnaire were declared valid.The reliability test of the questionnaire for knowledge and diet behavior was declared reliable with result 0,877 and 0,858. Conclusions: The validity and reliability test of the instrument have proven that the instrument of diet knowledge and behavior has good validity and reliability values.


Introduction
The capability of teenagers in behaving related to the physical changes that occur in the teenage's phase is that teenagers are able to accept physical states and make effective use of their body conditions.Howeverin this phase of development, it often causes its own problems so that teenagers are actually unable to accept the physical state due to the physical changes that occur. Weight problem is one of the main problems faced by teenagers, especially teenage girls. Paying attention to changes in body shape to look ideal is a form of teenagers response caused by changes in body shape due to weight problem 1 . The efforts made by teenagers to control weight are by doing a diet 2 .
Diet behavior is an individual activity in the regulation of dietary habit, drinking, and physical activity based on stimuli from the surrounding environment to lose weight. Healthy dietbehaviors are carried out appropriately, while unhealthy diet behaviors are carried out excessively 3 . Unhealthy diet behaviors in teenagers such as eating very few portions, skipping meals or taking diet pills are the starting point for eating disorders that cause concern for teenagers health 4 . Nutritional status is one way to determine the health status of teenagers 5 . Riskesdas data (2018) shows the prevalence of nutritional status of teenage girls aged 16-18 years in Indonesia, including very thin (0.5%), thin (3.8%), normal (79.8%), obese (11.4%), and obese (4.5%). The selection of diet methods carried out by teenage girls can be influenced by the teenage girl's knowledge of the diet 6 .
Knowledge is an important component in the formation of one's attitudes and behaviors 7 .
Nutritional knowledge is the result of a person knowing and understanding about things related to nutrition 8 . The level of knowledge and behavior of a person can be known by taking measurements using standardized measuring instruments to ensure the accuracy of the data and the consistency of the measurements collected. Standardized measuring instrumentsmust meet therequirements of validity and reliability 9 . A questionnaire is a set of structured lists of questions or statements that are used as a measuring tool to obtain information from respondents regarding perceptions, aspirations, behaviors, attitudes, circumstances, personal opinions, or other things according to the research or survey conducted 10 . A good questionnaire instrument will produce valid and reliable research data.So it is necessary to test the validity and reliability of the questionnaire first before the questionnaire is given to respondents. Invalid and reliable instruments will produce invalid and unreliable data when used to collect data 11 . Based on this description, this study aims to assess the validity and reliability of the questionnaire of knowledge and diet behavior in teenage girls.

Materials and Methods
This study is a descriptive examiner by examining the validity and reliability of the weight loss diet knowledge and behavior questionnaire in teenage girls that condected at SMA Negeri 1 Candimulyo. This research is quantitative, because the interpretation of research results is based on the results of statistical processing usingsoftware on a computer.The samples needed in this study were teenage girls aged 16-18 years who had been or were on a weight loss diet. The number of samples in this study was 30 teenage girls based on the minimum requirements for parametic statistical tests. Sample selection is carried out bypurposive sampling method.

Analysis of Difficulty Levels
Difficulty level analysis is an analysis carried out to find out whether the question is included in the category of easy or difficult questions. The level of difficulty is a number that indicates the difficulty or ease of a question 12  The criteria for the difficulty level index according to Arikunto (1999) are as follows 13 :

Differentiating Power Analysis
The differentiating power of the question is the ability of the question item in distinguishing students who have high abilities from students who have low abilities 13  The limits of the distinguishing power index criteria are as follows:

Distractor Analysis
An instrument used to measure learning achievement with multiplechoice question itemswith several alternative answers. The question item can be said to be good if the deceiver can function properly, which is at least chosen by 5% of respondents 13 .

Validity Test
Validation is a test to show how well the data collected from the research instrument 14 . Reliability is testing on research instruments that are carried out to find out how well the instrument provides stable and consistent results. The validity test of the diet behavior questionnaire was carried out usingLawshe's Content Validity Ratio (CVR) analysis. CVR analysis is a test performed to measure the validity of the contents. The CVR value is considered to be the higher the validity of its contents if the > 0 15 .

Discrimination Item Test
The power of item discrimination is a test carried out to measure the ability of items to distinguish between individuals or groups of individuals who have attributes and do not have attributes that are measured. Items that reach a correlation coefficient of ≥ 0.30 are included in the criteria for high discrimination power, while items with a correlation coefficient of < 0.3 are included in the criteria of low discrimination power 16 . The item discrimination power test was performed using the Pearson Product Moment correlation coefficient which was analyzed using SPSS version 25 software.

Reliability Test
Reliability tests are important because they refer to the consistency of all instruments 14 . The reliability test used in this study was the Cronbach's Alpha coefficient. The reliability test results are considered good if they have aCronbach's Alphavalue ≥ 0.6 or r alpha >r critical . The instrument is reliable when the value of Cronbach's Alphawhich is < 0.5 means low, 0.5-0.7 means medium, 0.7-0.9 means high, and > 0.9 means very good 17 .

Results and Discussion
The research was conducted by preparing a list of questions and questionnaire statements to be used. The questionnaire used in this study consisted of 25 questions regarding knowledge about the nutritional needs of teenagerswith multiple choice answer and 15 statements regarding teenagers dietbehavior in losing weight body with the answer using Skala Likert.Furthermore, testing was carried out on 30 girl students at SMANegeri 1 CandimulyoMagelang. The validity of knowledge questionnaire using analysis of difficulty levels, differentiating power, and distractor analysis. The behavioral questionnaire validity test uses an item discrimination test and Lawshe's Content Validity Ratio (CVR) analysis. While reliability testing using Cronbach's Alpha Coefficient test. The following are the results and discussions on eachquestionnaire test conducted:    Table 4 shows an analysis of the differentiating power of items on knowledge questionnaires as many as 1 item (3.3%) including criteria to be eliminated, 3 items (10%) including bad criteria, 9 items (30%) including the criteria are enough, 11 items (36.7%) include good criteria, and 1 item (3.3%) includes criteria for very well.  Table 5 shows the calculation of the distractor analysis on the knowledgequestionnairewhich is carried out based onnumber of answer choices which are then multiplied by the number of question items, but the key answer is notincluded in the calculation so that the total number ofdistractor is100 distractors.A total of 68 distractor are included in the good criteria, while 32distractor are included in the bad criteria.     Table 7 shows the results of the item discrimination test on the diet behavior questionnaire of 15question items, as many as 14 items (93.3%) were declared to have a high degree of discrimination because of the coefficient correlation value greater than 0.30 and 1 item (6.7) questions were stated to have low discrimination power because of the coefficient correlation value less than 0.30.

Conclusion
The validity test of the teenager nutritional needs knowledge questionnairethere were 24 items of questions declared valid. On the validity test of the dietbehavior questionnaire, all items of statements with a total of 15 statements were declared valid.Meanwhile in the reliability test,the two questionnaires were declared reliable or consistent with the value ofCronbach's Alphaknowledge questionnaire, which was 0.877 and the value of Cronbach's Alphaquestionnaire on dietbehavior was 0.858. The results of the trial of the validity and reliability of the instrument have proven that the research instrument on knowledge and diet behavior has a validity and reliability value that meets the criteria for used as a measuring tool for the level of knowledge ofteenagersnutritional needs and diet behavior in teenage girls.

Conflict of Interest
There is no conflict of interest regarding the publication of this paper.

Ethical Clearance
This article has been derived from a research projectapproved by Sebelas Maret University of Medical Sciences,with protocol number 01/02/09/2022/114

Authors' Contribution
Diva Amalia conceptualized and designed the study, prepared the draft of the manuscript and reviewed the manuscript. Tri RejekiAndayani conducted the study, data analysis and interpretation, assisted in drafting of the manuscript, reviewed the manuscript. SapjaAnantanyuasisted in drafting of the manuscript and reviewed the manuscript.