Levels of data week 1

• Ratio, Interval, Ordinal, Nominal are the four levels of measurement/data that form a hierarchy and determine what statistical analysis can be performed on the data.
• Different kinds of data are used in health psychology research, including participant’s weight, height, age, sex, depression scores, diet measured in calories, nutrients, high/low carb, protein, etc., and frequency of exercise.
• A popular classification of data types includes nominal, ordinal, interval, ratio, and categorical.
• Levels of data and their properties are important in health psychology research.
• Nominal data is distinguished by categories, such as ethnicity.
• Ordinal data is distinguished by order, such as small, medium, and large.
• Interval data consists of consistent intervals between numerical values, centigrade.
• Ratio data consists of consistent intervals between numerical values and a true, or absolute, zero, such as Kelvin.
• Individual responses to Likert items are often added together to provide a numerical total, and they are treated as interval rather than ordinal data.
• Interval data is where there is order and the difference (interval) between two values is meaningful.
• Interval data has a known and equal distance between each value on the scale.
• When the measurement equals zero, there is none of that variable.
• Ratio variables, on the other hand, never fall below zero, for example, weight in kilograms.
• Ratio scales hold no true zero and can represent values below zero.
• Most researchers agree that individual Likert questions/items are at the ordinal rather than interval level.
• For example, you can measure temperatures below 0 degrees Celsius, such as -10 degrees.
• The difference between 5°C and 10°C is the same as that between 2°5°C and 30°C – the difference between values can be compared.
• Examples of ratio variables include: enzyme activity, dose amount, reaction rate, flow rate, concentration, pulse, weight, length.
• Examples of interval variables/data include: Temperature (Fahrenheit), Temperature (Celcius), pH, SAT score (200-800), credit score (300-850), BDI, Extroversion, Intelligence.
• If something weighs zero kilograms, it truly weighs nothing.
• A ratio variable, has all the properties of an interval variable, and also has a true zero.
• Within Psychology Likert scales are often employed, for example, To what extent do you agree with the following statement …
• Reliability is the extent to which findings or measures can be repeated with similar results; consistency of measures.
• Validity is the extent to which instruments measure what they were intended to measure.
• Statistical conclusion validity is the adequacy of our operational definitions; do they really measure the variable of interest?
• Construct validity is the extent to which we can attribute our results to the independent variable; have we ruled out any confounds or alternative explanations?
• Internal validity is whether our findings generalise to other people; may be other people in the same population, or different populations.
• External validity is the extent to which results generalise to other settings; can sometimes be difficult to generalise from the lab to the 'real world'.
• Lab studies can be ecologically valid, for example, Milgram’s (1963) study was replicated in an office environment.
• Field studies can have low levels of ecological validity, for example, studies suggesting that nurses followed doctors orders to administer potentially fatal doses of medication (Hofling et al., 1966) have not been successfully replicated (Rank & Jacobson, 1977).
• Ecological validity is the extent to which findings generalise to other time periods; threats include research involving social factors being embedded in an historical context, for example, some argue that Asch’s (1956) conformity studies took place in a more conservative time when conformity was valued more highly.
• Time validity is the extent to which findings generalise to other time periods; threats include research involving social factors being embedded in an historical context.