Validity

          Validity with respect to the accuracy of measuring instruments to the concepts measured, so that only too well what it should measure at ¬ measure. For example, want to measure students' ability ¬ lam da math. Then the given problem with long sentences and complicated, so complicated elusive meaning. Finally students are not able to answer, due to not understanding the question. Another example, researchers wanted to measure the ability to speak ¬ way, but asked about grammar or literature such as poetry or rhymes. Gauge is not accurate (valid). Validity is not universally valid because depending on the circumstances and objectives of his research has ¬. Instruments that have been valid for a particular purpose is not automatically valid for other destinations.Examples of learning achievement and motivation variables can be measured by tests or by questionnaire. The trick can also be different, can test ¬ sanakan relief can be written or oral. There are three types of validity that is often used in the preparation of instruments, namely the validity of the contents, validity build understanding and validity of the forecast.

(a) The validity of the content 
           Content validity with respect to the instruments ability to measure the content of which must be measured. This means that the instruments capable of revealing the contents of a concept or variable to be measured. For example, achievement test social studies subject areas, should be able to reveal the contents of the field study. This can be done by preparing a test derived from the curriculum subject areas to be measured. In addition to the curriculum can also be enriched by seeing / reviewing the book source. Nevertheless achievement test is not possible to disclose all material existing in the field of study tar ¬ of course even if only for one semester. Therefore be taken in part from material in the form of the test sample. As a sample, so it should be able to reflect the material from all subject material. How the test sample taken in defining the concept is to choose the essential concepts of the material in it. For example set a number of concepts of every subject that exists. Of every concept developed at ¬ ing some test questions (see chart). This is where the pen ¬ importance of the role of lattice grid as a tool to satisfy content validity.

(b) The validity build understanding (Construct validity) 
           Validity wake up or wake up understanding (Construct validity) sustain ¬ tion with the ability to measure gauge sense definition contained in the material measured. Pe ¬ sense notion embodied in the concept of ability, interest, as a research variable in different fields of study must be clear what to measure. The concept is still an abstract concept, ¬ trust requires a more specific description, so easily measured. This means that each concept should be developed indikatomya indicators. With the indicator of each concept will get a sense of ease in the visible and specifies how measurement. For certain variables, made possible the use of measuring devices that measure a wide range of different ways.Establish indicators of a concept can be done in two ways, namely (a) use logical thinking or understanding on the basis of the theory of scientific knowledge and (b) use of empirical experience, ie what happens in real life.Example: The concept of "Social Relations", judging from experience, is the relevance of empirical indicators- Can get along with others- Groove or a lot of his friends- Accept others' opinions- Not impose their opinions- Can work together with anyone- And others.Measuring these indicators, means measuring the build understanding contained in the concept of social relations. Another example: The concept of attitude can be seen from the indicator in theory (deductive theory), among others, linkages of- Willingness to receive a stimulus object attitude- Willingness reacts to a stimulus object attitude- Assess the attitude object stimuli- Arrange / organize the attitude object- Internalization of values that exist in the object attitude.If test results show that the test indicators are not positively associated with each other, meaning that size has no validity build understanding. On that basis the indicator needs to be reviewed or corrected return. Another way to determine the validity of the wake is the notion of a gauge link (correlation) between the measuring instruments are made with a standard measuring instrument / standardized, if there has been standardized. When pointed ¬ a high correlation coefficient, the measure meets its validity.

(c) The validity of predictions (predictive validity) 
            Validity of prediction means is associated with certain criteria. In this validity test is preferred but not the content of the criteria, whether the instruments can be used to predict a trait or a certain behavior or certain criteria as desired. Suppose ¬ gauge learning motivation, whether it can be used to predict academic achievement achieved. This means that there is a positive relationship between achievement motivation. In other words it contains characteristics of validity and relevance of constancy or assessments (reliability). Motivation can be used to predict achievement scores when the scores obtained from the size of motivation are positively correlated with achievement scores. The validity of this prediction has two meanings. First the validity of short-term and second term. The validity of the short term, meaning that the instruments predictive power only for the period not longer. That is, scores are correlated at the same time. For example, assessments (reliability) occurred in the second semester means applies predictive power in the second semester, and not necessarily tar ¬ so in the next semester. While long-term validity implies score will be correlated also in the future. Given the validity of this greater emphasis on the correlation, the factors concerning the correlation dongan requirements must be met. These factors include the relationship of the concepts and variables can be explained on the basis of scientific knowledge, the mini ¬ mal sense and not making any sense. Another factor is the score that correlated to meet the linearity. Third validity ¬ explained above should ideally be used in preparing the instrument pe ¬ research has, at least two of validity, the validity of the content and validity of the ba ¬ built understanding. The validity of the content and build understanding ¬ absolutely necessary and can be pursued without testing when statically