Measurement in CSD Flashcards
Measurment
process, systematically assigning numbers, rules prescribed and understood by field, measuring attributes or features, goal in CSD- measurement of speech, language, and hearing variables, clear and practical set of rules (transparent)
Units of measurement
Data
Process
Identify, collect, organize, analyze, interpret, statistics (overlap with this- need this working knowledge of this)
Instrumentation Measures
prhysical variables- usually measured through an instrument, ex. leangth, time, form and frequency, air flow, tongue strength, acoustical measures
Observer Measures
behavioral measures (ex. language scores, mean length of utterance,
Classification of Data: Stevens Taxonomy
1) Nominal-level data
2) Ordinal-level data
3) Interval-level Data
4) Ratio-level Data
Nominal-level data
attributes of objects or events categorized, mutually exclusive categories, named groupings, only mathematical property: identity
ex: types of HL, gender
no hierarchy but random assignment- pass or fail.
what mathematical operation can be completed within this level?- can identify frequency of occurence,
Ordinal-level Data
Mathematical property of inequality (helps to rank order diff. variables), magnitude of data and identity, rank order (from least to most/most to least, greater or lesser value is assessed), mutually exclusive categories, difference between categories (ex. rating scale, Ranchos Amigos)
Interval Level Data
Mutually exclusive categories, named groups, equal intervals btw adjacent categories, rank ordered, most commpn example is Celsius or Fahrenheir scale) equal interval distance, no true zero (ex. same difference btw 2 adjacent categories, standardized scores in lang. tests - interpretive data)
Ratio-level Data
Mutually exclusive categories/named groups, equal intervals btw adjacent categories/constant distance btw adjacent intervals, rank ordered, TRUE ZERO, (ex. # of disfluencies, vowel duration, @ of misarticulations, speech intelligibility score)
Characteristics of Data
greater manipulation of data: more powerful statistical anaysis (converting data from 1 type to another), categorization important for statistical analysis, guidance of statistician,
histogram
frequency tables