Validity and reliability are absolute requirements that must be met when we conduct research using primary data (collecting data directly from respondents). Questionnaires or instruments used usually use questions with a Likert scale. Validity and reliability describe the extent to which the accuracy and precision of an instrument in measuring a phenomenon. Valid and reliable instruments have high accuracy and precision in measuring a phenomenon.

Validity refers to whether a measurement reflects what it is supposed to measure. Validity is a standard or basic measure that shows appropriateness, usefulness and validity that leads to the accuracy of interpretation of an evaluation procedure in accordance with the measurement objectives.

Kinds of validity are as follows:

Content Validity
Content validity describes the extent to which questions, assignments or items in an instrument/questionnaire can represent the overall and proportional behavior of the sample being tested. Content validity measures the degree of the test's ability to measure which includes the substance of the element to be measured. Content validity is used to measure learning ability, learning outcomes or learning achievement.

Construct Validity
Construct validity or Construct Validity is a validity that is concerned with how far the test items can measure what is really about to be measured in accordance with a specific concept or conceptual definition that has been set. Construct validity relates to events and objects that are abstract, but their symptoms can be observed and measured. Construct validity can be used to measure attitudes, self-concept interest, control focus, leadership style, achievement motivation and others, or maximum performance such as instruments to measure talent (aptitude test), intelligence (intellectual intelligence), intelligence, emotional and others.

Criteria Validity
Criterion validity or empirical validity (Criterion-Related Validity) is determined by criteria, either internal criteria or external criteria. The validity of the criteria is obtained through the results of test trials to respondents who are equivalent to the respondents who will be evaluated or studied.

Criterion validity is a measure of validity which is determined by comparing test scores with certain performances on external or other measures. An example of the use of criterion validity is the intelligence test which is correlated with the average academic score. Assuming, if a person's intelligence is high, then he will get good academic grades.

Advance Validity
Face validity is the type of validity with the lowest significance because it is only based on a cursory assessment of the content of the measuring instrument. If the content of the measuring instrument looks in accordance with what you want to measure, it can be said that advance validity has been met. Advance validity is also known as low validity from content validity.

Some principles when conducting validity tests, which include:

• The interpretation given to the assessment is valid only to the degree that it is directed to an evidence that supports its suitability and correctness.
• The use that can be made of the results of the assessment is valid only to the degree that it leads to evidence that supports its suitability and correctness.
• The interpretation and use of the assessment results is only valid when the values ​​obtained are appropriate
• The interpretation and usefulness of the assessment results are only valid when the consequences of this interpretation and usefulness are consistent with the value of concordance.

How to Calculate Validity
In testing the validity of each item, the score on the item in question is correlated with the total score. The score of each item is expressed by a score of X and the total score is expressed by a score of Y. By obtaining the validity index of each item, it can be seen which items meet the requirements can be seen from the validity index.

To determine the correlation coefficient between the test scores to be tested for validity and standardized test results owned by the same person, the moment product correlation formula can be used. Calculate the validity coefficient of the tested instrument (r-count), which has the same value as the correlation of the results of the previous step multiplied by the standardized instrument validity coefficient. Then compare the value of the validity coefficient with the value of the Pearson coefficient or Pearson table (r-table) at the significance level of a (generally chosen 0.05) and n = the number of appropriate data.

The criteria are as follows:

The instrument is valid, if r-count = r-table and the instrument is invalid if r-count < r-table. Determine the category of instrument validity whose reference is to the classification of validity proposed by Guilford:

• 0.80 < rxy < 1.00 very good validity (very high)
• 0.60 < ryx < 0.80 good validity (high)
• 0.40 < rxy < 0.60 sufficient validity (medium)
• 0.20 < rxy < 0.40 less validity (low)
• 0.00 < rxy < 0.20 poor validity (very low)
• rxy = 0.00 invalid

Reliability refers to how consistent an instrument is in measuring a phenomenon. If a reliable instrument is used to measure the same phenomenon repeatedly,
will produce scores that are not much different.

• Sugiono (2005). The definition of reliability according to Sugiono is a series of measurements or a series of measuring instruments that have consistency if the measurements made with the measuring instrument are repeated. Test reliability, which is the level of consistency of a test, is the extent to which the test can be trusted to produce a consistent score, relatively unchanged even though it is tested in different situations.
• Nursalam (2003). The definition of reliability according to Nursalam is the similarity of the results of measurements or observations if the facts or facts of life were measured or observed repeatedly at different times.
• Sukadji (2000). Understanding the reliability of a test according to Sukadji is the degree to which the test consistently measures the measured target. Reliability is expressed in the form of numbers, usually coefficients.
• Gronlund and Linn (1990). Understanding reliability according to Gronlund and Linn is the accuracy of the results obtained from a measurement.
• Sugiono (2005) in Suharto (2009). The definition of reliability according to Sugiono is a series of measurements or a series of measuring instruments that have consistency when the measurements made with the measuring instrument are repeated.
• Anastasia and Susana (1997). The definition of reliability according to Anastasia and Susana is something that refers to the consistency of scores achieved by the same person when they are tested with the same test on different occasions, or with a different set of equivalent items, or under different test conditions. different
• Suryabrata (2004). The definition of reliability according to Suryabrata is the extent to which the results of measurements with these tools can be trusted.