Scribd
Upload
2 views25 pages

Stat Encoding

The document outlines the process of data encoding, particularly focusing on survey data, and highlights the importance of data processing for analysis and interpretation. It provides common data processing methods and general encoding rules, as well as guidelines for special cases such as off-options responses, dichotomous variables, multiple responses, 'others' responses, and ordinal variables. The document emphasizes that encoding rules may vary among researchers and that data should be encoded according to the specific requirements of the study.

Uploaded by

๖ۣۜƊestR๏ψeR
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
2 views25 pages

Stat Encoding

The document outlines the process of data encoding, particularly focusing on survey data, and highlights the importance of data processing for analysis and interpretation. It provides common data processing methods and general encoding rules, as well as guidelines for special cases such as off-options responses, dichotomous variables, multiple responses, 'others' responses, and ordinal variables. The document emphasizes that encoding rules may vary among researchers and that data should be encoded according to the specific requirements of the study.

Uploaded by

๖ۣۜƊestR๏ψeR
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Data Encoding

• Many slides here are tackled in research classes, instead of


statistics classes.

• Some researchers use already-existing survey data set (or any


secondary data set from a valid research method) – without
doing their own encoding
Data processing aims to transform data into a version which may
be easily analyzed and interpreted

Analysis and interpretation becomes tedious without data


processing
Data processing methods vary depending on the data collection
method used
Common data processing methods
Data gathering Data processing Requirement
method method
Survey Encoding Spreadsheets
Observation Notetaking Anything that tracks observed
phenomena
Experiment Encoding/Notetaking Anything that records results of
/Recording experiment
Focus Group Transcription Voice recorder
We focus on encoding of survey data
Note:

The following encoding rules and guidelines are not universal

Some researchers have more rules or less rules than these


General rules
• Assign a row to every sample unit

• Assign a column to every variable

• Assign a cell to every value of the variable per


sample unit
Not all responses are encoded as is

There are cases which use a special set of codes


Some special cases in encoding
• I. Off-options response

• II. Variable has only two values / Dichotomous

• III. Multiple response

• IV. Presence of “Others” responses


– Probably the most problematic

• V. Ordinal variables
I. OFF-OPTIONS RESPONSE
• MISSING
• DON’T KNOW
• INVALID
• NOT APPLICABLE
• REFUSED
Common guidelines for such responses
• MISSING: Leave the cell blank

• DON’T KNOW/INVALID/NOT APPLICABLE/REFUSED: Assign a


special code
– Could be any character (preferably not solely numeric) that is rarely
used for the encoding
• For instance, 777, 999, DK, INV, NA, etc.
II. DICHOTOMY
– Yes/No questions
– Check/No Check questions

– Generally, questions with only two possible answers


Common guideline for dichotomy
• The 1,0 encoding
– YES/NO: Encode YES as 1, and NO as 0
– CHECK/NO CHECK: Encode CHECKED as 1, and NOT CHECKED as 0

– Generally, for dichotomous questions, name the column as one of


the two possible responses. Encode 1 if respondent possesses it, and
0 otherwise
III. MULTIPLE RESPONSE
Common guideline for single and multiple
• Provide one column for every single-response question

• Provide as many columns as the number of indicated options


for every multiple-response question
IV. “OTHERS” RESPONSES
Common guideline for “Others”
• Produce a column for every “Others” response that arises
during encoding
– Encode the values in this column as 1,0
V. ORDINAL VARIABLES
• Ordinal variables may be encoded with counting numbers, with
the highest number assigned to the highest value
– HOWEVER, such numbers may not be interpreted as usual numbers.
• They may only be arranged from lowest to highest or vice versa
BUT WHAT IF NONE OF THE 5 SPECIAL CASES IS
PRESENT?
Encode as is ☺
Data Encoding

You might also like