Choosing a research tool or instrum...

Choosing a research tool or instrument is single of the most important degrees in planning a research cogitation Research instruments must be pick outed or developed carefully to fit the research design and the plan for data analysis thus that the data collected will be useful for answering the research questions. righteous research instruments produce valid and reliable outcomes Validity and reliability often are caned psychometric properties of the research instrument, which means they show how well instruments measure the variables of interest to the researcher. nurtures who plan research studies sometimes picked research tools primarily because they are readily available or wen known; other give suck tos develop research instruments without adequate testing of their psychometric properties. in what way do you know that a research tool will bring into being meaningful research results?

MEASUREMENT CONCEPTS

Measurement is the proces of assigning numbers to give an account of the amount of a variable (ie, a characteristic, attribute, trait at hand in a person, object, situation subordinate to study). Nurses often assume that numbers portray by action the exact amount of the variable measured. In fact, the measurement ensue is an observed score. each observed score is made up of pair parts: a true score and an error score. The conformable to fact score represents the actual amount of the variable ready The difference between the observ score and the stanch score is due to measurement error; the amount of error cannot be measured directly. The smaller the error score, the more closely the observ score contemplates the true score or the actual amount of the variable at hand Measurement results that contain little error are said to be reliable.

Sources of measurement error include

* the instrument (eg improper calibration, variations in use),

* the environment (eg swing temperature, noise level),

* the researcher (eg fatigue, mood) and

* data processing (eg striking

the bad key while entering data into the computer mathematical errors in summing scores). Using measurement strategies that contract error increases the accuracy of measurement.

VALIDITY

Researchers make inferences from measurement terminates about how much of the variable being measured is near Validity refers to the compass to which these inferences are perfect A researcher's interpretation of a score is valid if it yields accurate conclusions about the variable. Validity, therefore, is not a characteristic of the research instrument itself, the word refers to the ways a researcher interprets and uses measurement results

Until lately validity was classified into three types: content-related, criterion-related, and construct-related. Today, validity is viewed as a single universal with construct validity as the central theme.2 form validity refers to the quality to which scores obtained from the use of an instrument are related to the universal of interest to the researcher. For example, an instrument designed to measure preoperative anxiety should be able to differentiate between that specific emotional state and a more persistent, general proclivity toward worry or uneasiness (ie, trait anxiety). To evaluate the soundnes of long heads about measurement results, the researcher must gather validity evidence. Estimates of invent validity may be based in succession different types of evidence, including content-related evidence and criterion-related evidence.

Content-related evidence supports the stretch to which the instrument measures all relevant dimensions of the fabricate often referred to as the universe or domain of the originate Such evidence may be obtained from a thorough search of the literature, from representatives of a relevant population, or from ease experts who are asked to arbitrator the extent to which items onward the instrument represent the defined easy in mind domain. Content validity also includes the readability of an instrument. There should be evidence that the instrument's on a level of readability matches the reading of the same height of the sample for which it is to be used.(3)

Criterion-related evidence supports the relationship between scores in succession the research instrument and another measure, known as the criterion. Criterion-related evidence is important if the researcher wants to make inferences from measurement be deriveds to some future behavior (ie, predictive validity evidence) or to a now passing behavior (ie, concurrent validity evidence). well adapted criterion-related evidence requires that the criterion measures themselves are valid, reliable, relevant, and without bias.(4) Criterion-related validity evidence usually is represented in terms of a correlation coefficient, an index whose values range from - 100 within .00 to + 1.00.(5) A correlation of 00 indicates no relationship between the measures. Negative values indicate that a high score in succession one measure is related to a gentle score on the other, and positive values mean that a high score upon one measure is related to a high score upon the other. The strength of the relationship, whether positive or negative, is indicated by dint of the actual value of the number. The higher the number, the stronger the relationship.