Research Design and Data

Link between study design and data, levels of measurement, data types with psychological examples

Which concepts are involved in measurement?

• Theoretical -> construct (abstract ideas e.g intelligence) -> operationalised 

• Measurement/design (e.g IQ test) -> must be reliable + valid 

• Data (links to variables + errors) -> observed effect from study (e.g % on IQ test) 

What are the broad categories for data types?

• Categorical = discrete number of response options (coded as integers) 

• Numeric = continuous variables which can take any real number value within the specified range of measurement 

What are the categorical data types?

• Nominal = binary/categorical variable where numerical markers share no relationship (no meaningful ordering) 

• Ordinal = binary/categorical variable where there’s a meaningful way to rank/order responses (BUT cannot meaningfully quantify the difference e.g likert scale) 

• Binary (special case) → only 2 possibilities 

What are the numeric data types?

• Interval/ratio - numerical values have meaning 

• Continuous 

• Discrete

  • Count = variables which can only take non-negative integer values (0, 1, etc)

Who coined the term ‘levels of measurement’?

Stevens (1946)

What are the levels of measurement?

• Interval 

  • No true 0 point on scale (0 does not mean absence of x variable) 

  • BUT can consider differences (which have a true 0 point) 

  • E.g IQ scores (debated) 

• Ratio 

  • Has true 0 point on scale (absolute 0, e.g can double/half values) 

  • So -> is plausible to multiple + divide ratio variables 

  • Can legitimately talk about double x 

  • E.g reaction time

• Nominal

  • Binary/categorical variable where numerical markers share no relationship (no meaningful ordering) 

• Ordinal

  • Binary/categorical variable where there’s a meaningful way to rank/order responses (BUT cannot meaningfully quantify the difference e.g likert scale) 

Data types and R

Data sets

What does ‘tidy data’ consist of?

• Each variable must have its own column

• Each observation must have its own row

• Each value must have its own cell

Each individual value belongs to both a variable + an observation