The most known and widely used types of reliability are: parallel forms, test-retest, interrater, and internal consistency reliability. Reliability is important because it gives precision, accuracy and dependable answers in researches. Consistency of measurement means there is a reliable measurement. If a method or particular piece of research is not reliable, researchers will and should lack confidence that the results obtained would be obtain again, and can thus hardly be confident that they reflect reality.
Even though that several examples of research, which reach similar results, even if the research is not strictly replicable, may still help researchers to increase their confidence in reliability, and perhaps also guide them toward appreciating the range of applicability of a theory (Kurpius & Stafford, 2006). Standard scores are the scores that are comparable because they are standardized in units of standard deviation while percentile or a percentile rank is a point in a distribution of scores below which a given percentage of scores fall.
It is a particular point within an entire distribution of scores. Percentile ranks have one major disadvantage in comparison to standard scores: percentile ranks have unequal units along their scale. Percentile ranks in the middle of the distribution tend to overemphasize differences between standard scores, where as percentile ranks at the tails of the distribution tend to underemphasize difference performance. The measure of central tendency is a statistical technique wherein a single number is used to represent a group of numbers.
Three different central tendency measures that are usually applied in researches are the mean, median and mode. These central tendency measures are significant because they have their own importance in measuring the central tendency. Mean provides the best estimate of average. However median is more reliable because it is clear and fixed on the other hand, mode is used when a quick and approximate measure of central tendency is required.