CHAPTER 3 RESEARCH METHODOLOGY
This chapter discussed the design and procedures undertaken during the conduct of the study. It presented the research method used, respondents of the study, instrument used, validation of instrument, data gathering procedures and statistical treatment of data.
Research Method Used The research method used in this study is the Descriptive Research. It is used to obtain information concerning the current status of phenomena to describe what exists with respect to variables or conditions in a situation. The method involved range from the survey which describes the status quo, the correlation study which investigates the relationship between the variable and the developmental studies which seek to determine changes over time (Key,1997). Descriptive research describes and interprets “what is”. It reveals conditions or relationships that exist or do not exist, practices that prevail or do not prevail, beliefs or points of view or attitudes that are held or are not held, processes that are going on or otherwise, effects that are being felt, or trends that are developing study (Salvador, Baysa & FuaGeronimo, 2008, pp. 58).
Descriptive method encompasses all the data gathered useful in adjusting or meeting the existing phenomenon.
The survey study was employed to
measure the existing event without inquiring into why it exists. As used in this research, gathered and treated are data on concessionaires profile such as concessionaire operations, number of years in business operations, operation time, monthly revenue, operating costs, amount allocated per month for operating costs. Respondents of the study The respondents of the study consisted of 30 concessionaires of Polytechnic University of the Philippines-Manila. The selection of the participants was done through random sampling using ___ technique. Out of the 30 concessionaires, ___ were from north wing and the remaining ___ were from east wing. Instrument Used In this research study, the researchers utilized a five-part Questionnaire. The first part aimed to gather data on the profile of the 30 respondents from the Polytechnic University of the Philippines in terms of concessionaire operations, number of years in business operations, operation time, monthly revenue, operating costs, and amount allocated per month for operating costs. The second part focused on internal factors of concessionaires. It consists of 4 categories. The first category is all about concessionaire’s background, consisting of 9 items. The second category talked about improve management of concessionaire, consisting of 9 items. The third category is concerned on financial status of
34
concessionaires, consisting of 10 items.
The fourth category talked about
accounting guidelines of concessionaire, consisting of 10 items. The third part concentrated on operations management best practices regarding to
financial
management, financial aspect, personnel and employees, health and sanitation, and customer care. The fourth part focused on financial practices in terms of general policies and procedures, source documents.
The fifth part included
problems encountered and proposed solutions. Likert scale was used to modify the answers of the respondents considering five (5) options with the corresponding scale. The options are (5) Strongly Agree; (4) Agree; (3) Undecided; (2) Disagree; and (1) Strongly Disagree. Respondents are requested to check the space provided for each option. The consolidated points from the respondents’ answers to each item over a five-point scale were as follows:
WEIGTED MEAN
SCALE
VERBAL INTERPRETATION
4.51 - 5.0
5
Strongly Agree
3.51 - 4.5
4
Agree
2.51 - 3.5
3
Undecided
1.51 - 2.5
2
Disagree
1.0 - 1.5
1
Strongly Disagree
Validation of the Instrument
35
Validity refers to the degree in which our test or other measuring device is truly measuring what it intends to measure. As to the determination of the questionnaires’ validity, the researchers used Item Analysis, by which the difficulty index (DI) and the validity index (VI) of each item in test were determined. This was used to determine if the respondents would be able to answer the instrument.
Eight (8) representatives of concessionaires from
University of the Philippines-Diliman who were not part of the respondents are the sample used. The responses of these individuals were tallied to determine who answered the questions with strongly agree and strongly disagree. As to difficulty index, a question was not appropriate if it is easy, appropriate if moderately difficult, and modified if difficult. The Difficulty Index formula is: (PSA – PSD) DI = 2 where: PSA = Percentage of Strongly Agree PSD = Percentage of Strongly Disagree The ranges of values used in determining the Difficulty Index are the following: DI ≥ .75
Item is Easy (E)
.25 ≤ DI ≤ .75
Item is Moderately Difficult (Mdf)
DI ≤ .25
Item is Difficult (Df)
The result of the difficulty index computation of the questionnaire shows that all were marked Moderately Difficult which shows that the questions are valid and appropriate. 36
On the other hand, questions that resulted to a validity index of highly discriminating were maintained and those that resulted to moderately discriminating were modified. The Validity Index formula is: VI = PSA– PSD To get:
PSA = NSA N PSD = NSD N
W here:
VI = Validity Index PSA = percentage of NSA PSD = percentage of NSD NSA = answer with highest frequency NSD = answer with lowest frequency N = number of sample respondents
The ranges of values used in determining the Validity Index are the following: Vi ≥ 0.75
Highly Discriminating (HD)
0.25 ≤ Vi ≤ 0.75
Moderately Discriminating (MD)
VI ≤ 0.25
Not Discriminating (ND)
The result of the validity index computation of the questionnaire shows that Highly Discriminating which shows that the questions are to be retained and Moderately Discriminating which shows that the questions are to be modified. The split-half method was used to test the reliability of the questionnaire. The total set of item questions was divided into halves. The odd-numbered items are assigned as variable x, while the even-numbered are assigned as variable y. The scores on the halves were correlated to obtain an estimate of reliability. 37
To obtain the value of the correlation, which is the split-half reliability, the Pearson r Product-Moment Correlation Coefficient formula was used. The Pearson r Product-Moment Correlation Coefficient formula is: n∑xy - ∑x∑y r= √[n∑x2 – (∑x)2 ][ n∑y2 – (∑y)2 ]
The formula resulted to 95.5% which denoted a very high correlation. Since the split-half procedure was based upon a correlation between scores obtained on only half the test, a correction was needed to determine the reliability of the entire test. The halves can be considered approximation to alternate forms. To do this, the Spearman – Brown Prophecy Formula was used. The Spearman-Brown Prophecy Formula is: 2r rsb = 1+r where: rsb = Reliability of the whole instrument r = Split-half Correlation The formula resulted to 97.7% which denoted very high correlation. The computation for validity and reliability are presented in the Appendix D of the study.
Data Gathering Procedures The researchers gathered information from the Internet, published books and unpublished theses to choose a topic for research. After deliberation and pieces of advice from the thesis adviser, the researchers finally had their topic. 38
The researchers began to collect information about the topic and started to construct their statement of the problem, conceptual framework, and sample questionnaire. These are submitted to their professor for improvement and verification. Soon, the researchers wrote a formal letter addressed to the concessionaires with whom they wish to forward the survey questionnaire and asked their adviser to sign. This letter assured the respondent’s anonymity of any information that was supplied. Also, it ascertained that everything was solely for academic purposes. With regards to gathering of data, ***********to be continued di pa tayo nasusurvey
Statistical Treatment of Data The researchers employed the following statistical technique and procedures to analyze and interpret data gathered from the respondents of the study. The data were coded, tallied and tabulated to facilitate the presentation and interpretation of results using the following:
1. Sampling Size This technique was employed to compute the size of the sample obtained from the total population. The formula for computing the size of the sample is as follows. n = where: 39
n = the size of the sample N = the size of the population e = margin of error 2. Frequency Distribution This was utilized to present the data clearly and accurately in tabulated form with corresponding roles and columns for each specified category.
3. Sample Percentage This technique was used to determine the proportion of one class relative to the total number of respondents in all classes. In particular, it was to analyze the specific items in the profile of the respondents. The formula for computing simple percentage is as follows: P = where: P = percentage F = frequency respondents belonging to class N = total number of respondents in a class 100 = constant multiplier
4. Frequency and Percentage The percentage and frequency distributions were used to classify the respondents according to personal background variables such as 40
concessionaire operations, number of years in business operations, operation time, monthly revenue, operating costs, and amount allocated per month for operating costs.. The frequency also presented the actual response of the respondents to a specific question or item in the questionnaire. On the other hand, the percentage of that item is computed by dividing it with the sample total number of respondents who participated in the survey. The formula used in the application of this technique is: % = (f/n) x 100 where: % = percentage f = frequency n = number of cases or total sample 5. Ranking This is a descriptive measure to describe numerical data in addition to percentage. Ranking was used in the study for comparative purpose and for sharing the importance of items analyzed. 6. Weighted Mean Another statistical technique used by the researchers was the weighted mean. It was used to determine the average responses of the different options provided in the various parts of the survey questionnaire used. The method is used in conjunction with the Likert Scale. It was solved by the formula: x=∑ x/n where: x
= weighted mean
41
∑ x = the sum of all the products of and x, the frequency of each weight of each operation n = total number of respondents
being
7. One-Way Analysis of Variance (ANOVA) A One-Way Analysis of Variance is a way to test the equality of three or more means at one time by using variances. This was used to test the significant differences between means of 3 or more sets of data simultaneously. It is the method of dividing the variation observed in experimental data into different parts, each part assignable to a known source of factor. ANOVA set at 0.05 probabilities was used. a. Sum of Squares within: SSt = ∑ X2t - (∑Xt)2 N where: SSt X2t (Xt)2 n
= total sum of squares = total sum of x2 = square of total sum of x = total number of samples
b. Sum of Square between: SSb = ( X– Xt)2 n where: SSb X Xt n
= sum of square between = mean of any group = mean of total distribution = total number of sample
c. Sum of Square total: SSt = SSb– SSw where: SSw = sum of square within SSt = sum of square total 42
SSb
= sum of square between
d. Mean Square: SSb MSb =
dfb
where: MSb = mean square between SSb = sum of square between dfb = degrees of freedom between
SSw MSw=
dfw
where: MSw = mean square within SSw = sum of square within dfw = degrees of freedom within e. To compute F: SSC -1 SSF N-K
F=
where: Tc nc Σx k N
= column total for all observations in the treatment = number of observations (sample size) for each respective treatment = sum of all observations = number of treatments = total number of observations
Computation for degree of freedom is necessary in order to obtain the critical value (P), the formula is: df where: r k df
= (r-1) (k-1) = number of rows = number of columns = degree of freedom
43
8. Likert Scale Statement of opinions is presented to respondents who indicate agreement or disagreement within the statement.
WEIGTED MEAN
SCALE
VERBAL INTERPRETATION
4.51 - 5.0
5
Strongly Agree
3.51 - 4.5
4
Agree
2.51 - 3.5
3
Undecided
1.51 - 2.5
2
Disagree
1.0 - 1.5
1
Strongly Disagree
44