Classroom Observation Data Collection Technique

79 8 In the preliminary testing, the representatives of the user were also asked to fill out the questionnaire in order to gain feedback from them. The blueprint of the users’ validation is attached in appendix 5. In the main field testing evaluation the questionnaires were given to the students related to the application that the students have tried. The form of the questionnaire was the Likert-scale model in order to know how well the application helps them to improve their speaking skill. The concept of the questionnaire was based on the theory for evaluating hypermedia by using the Likert-scale questionnaire in main field testing for the students is in appendix 5. Besides conducting the questionnaires, the data was obtained from the interview and classroom observation as well. The interview was conducted on October 10, 2015 to gather information from the English lecturer of Stikes Wira Husada related to the students’ perceived needs and real need, the gap of the students, their learning style, the method used teaching English, the technology used, and the students’ proficiency. The interview and classroom observation gave additional information that might not be captured in the questionnaires. The blueprint of the interview for need analysis is attached in appendix 6. The second interview was conducted to gain more feedbacks from the IT experts. The interviews were conducted twice, on the 12 April 2016 and 15 April 2016. The third interview for the first expert validation on the learning content was conducted on 10 May 2016. While the other two expert validators were interviewed on the 16 May 2015. The interview to gain feedback from the users was represented by five students of Stikes Wira Husada in which it was conducted on 11 May 2016. 80 8 Therefore, the interviews in the preliminary field testing were gained from the experts both for the learning content and the application and the users for the application and learning content as well. The guidelines for the interviews are attached in appendix 7. The data gathering instruments that were used beside questionnaire and interview was classroom observation in which the researcher observed the learning environment, the learners’ characteristics, the classroom activity, the techniques that the teacher used in speaking English, the media used for teaching English as well as the feedback used from the teacher. The researcher video-taped the observation and record a note related to the observation.

E. Data Analysis Technique

There were two questionnaires distributed to the students, the first one was used to provide the data for need analysis and the second questionnaire was used in o rder to discover the learners’ perception after using the Android application. The questionnaire for the need analysis was in form of questions related to the learning materials, teaching techniques, and the use of media in learning English especially in speaking. The data obtained from the need analysis questionnaire was in the form of close-ended. The close-ended form was counted statistically using SPSS V.23. The data from the need analysis then was used to developed the Android application to help the nursing students to learn speaking English, specially to determine the features which are suitable to support learning speaking. PLAGIAT MERUPAKAN TINDAKAN TIDAK TERPUJI 81 8 The second questionnaire which was intended to answer the second question of research problem was in the Likert-scale model. The questionnaire focuses on the students’ perception after they use the application. The data gathered from the questionnaire then analysed. The Likert-scale ranged from Strongly Disagree, Disagree, Not Sure, Agree and Strongly Agree Best Kahn, 2006. The scale value of each response is described in table 3. 3. Table 3.3. Scale value of Likert scale Statement Value Strongly Agree 5 Agree 4 Not sure 3 Disagree 2 Strongly Disagree 1 The next step is counted the answered questions of each question using SPSS V.23. In inputting the data in SPSS, the variable or the statement was giving in code. For instance, A1 is for statement 1 in need analysis questionnaire. After finish inputting the variables the data was entered based on the response given by filling in the value of each response. The result then analysed using descriptive frequencies which included the frequency of each statement, mean and standard deviation. The frequency of the response of each statement in forms of its mean is represented in table 3.4. and the complete data of each response for the result of need analysis, 82 8 expert validation on learning content, expert validation on the Android application and users’ validation is attached in appendix 13, 14, 15 and 16 respectively. Table 3.4. The Template of questionnaire result based on the mean and its category. STATEMENTS Code STATEMENTS N mean Category Valid missing In order to discover criteria of each statement in the questionnaire, Criterion Reference Evaluation CRE proposed by Sukarjo 2006 the result of the mean was classified. The formula for classifying the category as follows: Table 3.5. Criterion Reference Evaluation formula. Category Interval Very HighVery Good Very acceptable Xi + 1.80 SD X High Good Acceptable Xi + 0.60 SD X ≤ Xi + 1.80 SD Fair Xi – 0.60 SD X ≤ Xi + 0.60 SD Low Poor Unacceptable Xi – 1.80 SD X ≤ Xi – 0.60 SD Very Low Very Poor Very unacceptable X ≤ Xi – 1.80 SD 83 8 Notes: X = actual mean score Xi ideal mean score = ½ maximum ideal score + minimum ideal score SD standard deviation = 1 6 maximum ideal score – minimum ideal score The maximum ideal score is the highest score based on the scale value of the Likert scale, that is 5. While the minimum ideal score is the lowest scale value, that is 1. Therefore, the Xi ideal mean score is 3 and the Standard Deviation SD is 0.6. Five scale in scale value five equals six in standard deviation scale. Therefore, one scale in scale value 5 = 65 =1.2 standard deviation scale. In various sources 1.2 standard deviation scale is rounded to 1.0 standard deviation scale. Based on the CRE described in table 3.5 above, when it is related to the application product, then if the result score of each statement in the questionnaire is very high or very good, it means that it does not need revision. If the result score is high or good, then revision is optional. If the result score is fair, then the revision is needed. Next, when the result score is poor the mayor revision is needed for some parts of the application. Furthermore. When the score result is very poor the it is highly recommended to revise or change some parts of the application. In addition, related to the acceptability of the Android application for the users to learn speaking English, the mean score category ranges from very acceptable to very unacceptable. The Android application would be considered as acceptable for the users when it is able to represent the efficiency of a mobile learning related to PLAGIAT MERUPAKAN TINDAKAN TIDAK TERPUJI 84 8 its navigation pane, screen design, information presentation, media integration, and overall functionality. Besides the close-ended part of the questionnaires, the open-ended part was also needed to be chosen accordingly and organised based on the category. After that, the data was described and explained in order to support the quantitative data obtained from the close-ended part. The final step was interpretation of the data. Another obtained data was a qualitative data in forms of open- ended questions and interviews which were analysed and coded based on the blue print and interview guidelines. The interview transcription was chosen accordingly and presented in forms of quotation narrated in chapter IV in order to support the quantitative data.