TECHNIQUE OF ANALYSIS DATA AND TESTING OF HYPOTHESIS

TECHNIQUE OF ANALYSIS DATA AND TESTING OF HYPOTHESIS

       I.      TECHNIQUE OF ANALYSIS DATA

The process begins with data analysis predict all the data available from various sources, i.e. interviews, observations, which are written in field notes, personal papers, official documents, photographic images, and so forth. Once read, studied, and performed the data reduction by making abstraction of an effort to make a summary of the core, and the statements that need to be maintained so as to stay within it. The next step is arranged in units and categorized. The final stage of data analysis is examining the validity of the data.

The technique in data analysis include: basic concepts of data analysis, degeneration unit including the categorization of data validity checks, and then ends with the interpretation of data.

A.    Basic Concepts of Data analysis

According to Patton, 1980 (in Lexy J. Moleong 2002: 103) explains that data analysis is the process of ordering the data; organize it into a pattern, category, and the basic outline of the unit. Meanwhile, according to Taylor, (1975: 79) defines data analysis as a process that details a formal effort to find a theme and formulate hypotheses (ideas) as suggested and an attempt to provide assistance and the theme of the hypothesis. So the data analysis is the process of organizing and sorting data into patterns, categories and the basic outline of the unit in order to discover the theme and can be formulated as a working hypothesis based on the data.

The first step is to organize the data analysis of the data. The data collected may consist of field notes and researcher comments, images, photographs, documents, in the form of reports, biographies, articles, and so forth. Employment data in this analysis is to organize, sort, classify, provide the code, and reorganized. Organizing and managing the data is aimed at finding themes and working hypotheses which eventually became the substantive theory.

Work of analyzing the data requires concentration and effort exertion, the mind of researchers In addition to analyzing data. Researchers also need to be and still need to explore the literature in order to confirm the theory a new theory that might be found.

baca selanjutnya

B.     Processing Unit

A description of the degeneration of this unit consists of units and arranging tipelogi units.

1.      Tipelogi unit.

According to Lofland and Lofland, (! 984) (Lexy in 2002: 190), a unit of social life is the unanimity in which someone asks a question. Linciln and Guba (1985: 344) were named the unit as a unit of information that serves to determine or define the category.

Accordingly, Patton, (1987: 306-310) distinguishes two types of units: The original type, and construct the type of analysis. Patton stated that this is the original type using anthropological perspective. It is based on the assumption that the social and cultural behavior should be studied from the point of view of human behavior and definition. Thus, the conceptualization of the unit should be found by analyzing the cognitive processes of people who studied, not in terms of entnosentrisme researchers.

2.      Preparation of unit

Lincoln and Guba (1985: 345) say that the first rare degeneration is a unit of analysis should be carefully read and studied all kinds of data that is collected. After that, keep the units were identified. Researcher is entering into the index card. Preparation of units and inclusion into the index card should be understood by others.

C. Categorization

Categorization in the description consists of (1) the functions and principles of categorization, and (2) rare-step categorization described as follows.

1.      Functions and principles of categorization

Categorization means the preparation of the category. No other category is one of the stacks of a set of stacks that are prepared on the mind, intuition, opinion, or specific criteria. Linclon and Guba describe further categorization is (1) grouping the cards that have been made into parts of the contents of which are clearly related, (2) to formulate rules that describe the categories and that ultimately can be used to determine inclusion in the category of each card, and also as a basis for checking the validity of the data, and (3) ensure that each category has been drawn up with each other following the principles consistently.

2.      Categorization measures

The method used in the categorization is based on the method of comparative analysis of the steps outlined above ten rare, which is the last step is the analysis should predict the entire category once again not to get forgotten. After completion of the analysis, the author shall be held before interpreting validity examination of the data, the examination can be performed using the technique of data validity checks.

D.    validity of the data

To avoid mistakes or errors of data has been collected, is necessary to check the validity of data. Checking the validity of the data based on criteria degree trust (credibility) with triangulation techniques, perseverance observation, checking peers (Moleong, 2004).
Triangulation is a technique of checking the validity of the data that is based on something outside of the data for checking purposes or as a comparison against the existing data (Moleong, 200). Trigulasi used is the source, i.e. comparing observation outcome data, results of student work and interviews of the subjects outlined in the application of effective methods to aid reading.

                II.            TESTING OF HYPOTHESIS

Hypothesis is defined as a temporary response to the formulation of research problems. The truth of the hypothesis must be proved by the data collected. In statistical hypothesis is interpreted as a statement about the state of the population (parameter) that will be verifiable based on data obtained from the study sample. Therefore, the statistical hypothesis tested is zero. Null hypothesis (Ho) is a statement of no difference between the parameters of the sample. Opponents of the null hypothesis is the alternative hypothesis (Ha) stating the differences between the samples parameters.

A.    standard error

Basically the hypothesis test is to assess population parameters based on the data sample. There are two ways of estimating that,

1.      A point estimate is an estimate of population parameters based on an average value of sample data.

2.      Interval estimate is an estimate of population parameters based on the average interval sample data.

Assessing population parameters using a single value (point estimate) will have a higher risk of error compared to using the interval estimate. the greater the interval his estimates the smaller faults. For further estimation error is expressed in the form of percentage chances. Usually in this study are set in advance the estimated error, which is used is 5% and 1%.

B.     Two Errors in Hypothesis Testing

In assessing population parameters based on sample data, the possibility that there are two errors,

a.       Type I error is a mistake to reject Ho is true, which is expressed by α.

b.      Type II error is an error when accepting the hypothesis is wrong (it should be rejected). This error rate to certify with β.

When the sample data obtained from data collected with the same population parameter values​​, then the hypothesis is formulated 100% acceptable. So there are no errors. The error rate is called the level of significant. Usually the level of significance is taken as 1% and 5%.

C.     Kinds of Hypothesis Testing

There are three kinds of hypothesis testing, ie test the two parties, the right and left sides. This type of test depends on the sound of a sentence hypothesis.

a.       Two Tail Test

Filing Ho and Ha in two test directions are as follows:

Ø  Ho: written in the form of equation (using the sign =)

Ø  Ha: written by using the sign ≠

b.      Left Tail

Test is used when the left side: Ho read "greater than or equal to (≥)" and Ha read "smaller" (<), the word is less than or equal to synonym "smallest".

Example:

Ho: μ_1 ≥ μ_2

Ha: μ_1 <μ_2

c.       Right Tail

Right to use the test if Ho read "less than or equal to" (≤) and Ha read "larger" (>) or "biggest".

Example:
Ho: μ_1 ≤ μ_2

Ha: μ_1> μ_2