Term
|
Definition
|
|
Term
|
Definition
| set of individuals, objects, or scores interested in studying |
|
|
Term
|
Definition
| property or characteristic that can have different values at different times |
|
|
Term
|
Definition
| variable that is systematically manipulated by investigator |
|
|
Term
|
Definition
| variable that investigator measures to determine effect on independent variable |
|
|
Term
|
Definition
| measurements made on subjects |
|
|
Term
|
Definition
| number calculated on sample data that quantifies a characteristic of the sample |
|
|
Term
|
Definition
| number calculated on population data that quantifies a characteristic of the population |
|
|
Term
|
Definition
| Nominal, Ordinal, Interval, Ratio |
|
|
Term
|
Definition
Definition: classify or categorize people, places, or things; represents qualitative differences
Example: Race/Ethnicity, City
Operations: comparison with 2 possible outcomes X=Y or Xnot=Y |
|
|
Term
|
Definition
Definition: when people, places, or things can be rank ordered
Example: satisfaction, finishing order in a race
Sidenote: differences between levels not always same; 1st and 2nd can be close, 3rd and 4th further
Operations: comparisons X=Y, Xnot=Y, XY |
|
|
Term
|
Definition
Definition: people, places, things can be rank ordered, difference between adjacent value is same, #'s have constant magnitude, add & subtract, no meaningful 0
Example: temperature (F or C), Years
Operations: X-2=Y-2, X-2not=Y-5, X-1Y-6
Sidenote: ratio of differences is meaningful, differences of ratios is not meaningful |
|
|
Term
|
Definition
Definition: interval scale with meaningful zero
Sidenote: differences meaningful, ratios meaningful
Example: Age, Weight, Height, Elapsed Time
Operations: all comparisons (equality, inequality, less than, greater than) |
|
|
Term
| Delimited ASCII (text files) |
|
Definition
| each subject is row, each variable separated by delimiter (comma, tab, semicolon, etc) |
|
|
Term
| Formatted ASCII (text files) |
|
Definition
| each subject is row, each variable begins and ends in specific column |
|
|
Term
|
Definition
| specially formatted files for holding rows and columns of data; each column can have properties |
|
|
Term
|
Definition
| database management programs (Access, Oracle, SQL); data often in multiple tables which can be linked by key field |
|
|
Term
| Other Statistical Software Data Files |
|
Definition
| SAS, STATA, Old versions of SPSS |
|
|
Term
|
Definition
| reading data into a program from an electronic data source |
|
|
Term
|
Definition
| writing data into a format that can be read by other programs |
|
|
Term
|
Definition
theoretically can have infinite # of vlaues between adjacent units of scale
Ex: seconds, pounds, ounces of beer left in keg |
|
|
Term
|
Definition
there are no possible values between adjacent units on scale
Ex: # of cookies in package, # of kegs emptied, # of children in family |
|
|
Term
| Real Limits of Continuous Variable |
|
Definition
| values above and below recorded value by 1/2 of smallest unit of measurement |
|
|
Term
|
Definition
1 - identify missing values and set all value codes
2 - drop or eliminate cases with too much missing data
3 - think about each variable
4 - get data set properly cleaned before analysis |
|
|
Term
|
Definition
|
|
Term
|
Definition
| documents how data are coded |
|
|
Term
|
Definition
| used to exclude cases with bad data |
|
|
Term
|
Definition
| allow for writing equations |
|
|
Term
|
Definition
| If X then do Y; X is logical condition that is true or false; Y is action that changes data |
|
|
Term
| RECODE (discrete variables) |
|
Definition
| recoding into different variable and updated value codes |
|
|
Term
|
Definition
| faster; allows writing and saving series of SPSS commands; easier to make mistakes |
|
|
Term
|
Definition
|
|
Term
|
Definition
|
|
Term
|
Definition
| number of scores with this value or less |
|
|
Term
|
Definition
| percentage of scores with this value or less |
|
|
Term
|
Definition
| difference betweeen lowest and highest value in the data |
|
|
Term
|
Definition
| trade off between a few (less information) and many (overwhelming information) |
|
|
Term
|
Definition
|
|
Term
| Percentile or Percentile Point |
|
Definition
| value of the measurement scale below which a specified percentage of the scores in a distribution fall |
|
|
Term
|
Definition
| percentage of scores with values lower than the score in question |
|
|
Term
| 95th and 99th percentiles |
|
Definition
| often used to identity people whose performance or characteristics are unusual |
|
|
Term
|
Definition
| divided into 3rds below 33rd percentile, between 33rd and 67th, and 67th and above |
|
|
Term
|
Definition
| divided into 4 equal parts using 25th, 50th, and 75th percentiles |
|
|
Term
|
Definition
| divided into 5 equal parts using 20th, 40th, 60th, and 80th percentiles |
|
|
Term
|
Definition
| divided into 10 equal parts |
|
|
Term
|
Definition
| sum of scores divided by number of scores |
|
|
Term
| Estimate Mean from Grouped Frequency Distribution |
|
Definition
1 - multiply midpoint of each interval times frequency in each interval
2 - add up products
3 - divide the sum by the total N |
|
|
Term
|
Definition
| scale value below which 50% of scores fall; 50th percentile, centermost score if odd #, average of 2 centermost scores if even # |
|
|
Term
|
Definition
| most frequent score in distribution |
|
|
Term
|
Definition
| differnce between highest and lowest scores in distribution (Range = Highest - Lowest) |
|
|
Term
|
Definition
| differnce between highest and lowest scores in distribution (Range = Highest - Lowest) |
|
|
Term
|
Definition
|
|
Term
|
Definition
| square root of the varience |
|
|
Term
| Interquartile Range (IQR) |
|
Definition
| derived from scores that define the middle 50% of a distribution; computed as 75th percentile score - 25th percentile score |
|
|
Term
|
Definition
| how asymmetric distribution is; symmetrical - 0 skew; negative - tail on left; positive - tail on right |
|
|
Term
|
Definition
| how peaked distribution is; platykurtic (negative) - distribution is relatively flat; mesokurtic (0) distribution is relatively normal; leptokurtic (positive) relatively peaked |
|
|
Term
| Divide distribution into k intervals that are i units wide |
|
Definition
|
|
