Download as doc, pdf, or txt
Download as doc, pdf, or txt
You are on page 1of 23

[Type here]

PROCESSING THE DATA

 Prepare the data for processing,

 set up a system to oversee data processing, and perform data entering.

 Edit the data to make it cleaner so that it may be analysed.

 Create tabulations incorporating the indicators.

 Distribute and archive data


[Type here]

Overview

THIS IS BECAUSE THE MICS5 DATA-PROCESSING SYSTEM PROCESSES DATA IN REAL TIME

AS SURVEYS ARE BEING CONDUCTED IN THE FIELD, ALLOWING FOR EXTREMELY RAPID TASK

COMPLETION. AS SOON AS THE SURVEYS ARE RETURNED FROM THE FIELD, THE DATA FOR EACH

CLUSTER IS PROCESSED IN ITS OWN INDIVIDUAL DATA FILE. THE DATA CAN BE PROCESSED IN

STAGES WHILE THE FIELDWORK IS STILL ONGOING. THIS MEANS THAT THE VAST MAJORITY OF

THE DATA HAS BEEN PROCESSED BY THE MOMENT WHEN THE FINAL QUESTIONNAIRE ARE

COMPLETED AND SENT TO HEADQUARTERS.

TABLE 1
Page |3
[Type here]

PREPARATION FOR DATA ENTRY

One reason to get the data input phase ready ahead of time is so that fieldwork may get

underway without delay. During the planning stage, you'll be doing the things on this list:

First Stages of Data Processing

Primary data processing aims to provide revised, error-free data sets. Here are some of the

most important processes involved in data processing:

All surveys for a cluster must be entered onto a data file, and then all questionnaires for

that cluster must be entered into the data file.

Backing up the reviewed and validated data file, making secondary edits to the data file,

then backing up the updated data file after those edits have been made, are all examples of

data backup.

PREPARATION OF RESULTS FROM SECONDARY DATA

Production of analytical data files and the MICS5 standard tables are the two main

outcomes of secondary data processing. The steps of secondary data processing are as

follows:

We construct a single data file that consolidates all of the cluster's data files, and we fill in

any gaps. Procedures for transferring data to SPSS and determining appropriate sample

weights In order to facilitate analysis, data must be recoded, a wealth index must be built,

tables must be created, and data must be disseminated and stored.


PEOPLE AND FACILITIES

PERSONNEL

A questionnaire administrator, operators in charge of inputting the data, secondary

editors, and a supervisor all make up the data processing team for a MICS5 survey. Since

each role is responsible for a unique set of tasks, merging them is likely to compromise the

integrity of the information you collect.

The administrator of the questionnaires verifies and sorts them when they are returned

from the field (OR OFFICE EDITOR). A CLUSTER is responsible for ensuring that all of

the questionnaires are present and in order when they are brought into the data processing

office. Using the help of the field crew, he or she is responsible for locating any unanswered

questions (THE PRECISE STEPS THAT THE QUESTIONNAIRE ADMINISTRATOR

MUST TAKE ARE DETAILED LATER).

The Hardware That Makes Computers Work

Listed below is the hardware that will be required for data processing:

Methods of Data Entry

Computer networking and backup storage for the data processing manager's workstation

(for example, a portable USB device for operators to transfer files to the data-processing

supervisor, if network is not established) paper surge protection consisting of a printer,

toner cartridges, and printer ribbons

Page |5
[Type here]

Pens in blue for UPS (discontinuous power systems)

WORKING AREA

There can't be the same workspace for both data entering and data editing. There must be

enough desk space in the data entry area for each operator to set up their computer and

questionnaire. Working surfaces and electrical outlets must be in sufficient supply. The

space should be as bright, dry, and frigid as possible. This necessitates the use of air

conditioning in regions with excessive heat. without pause

As shown in Table 2, the many kinds of questionnaires and the characteristics that

distinguish them are listed.

Questionnaire Cluster number Household number Line number


Household HH1 HH2
Women HH1 HH2 LN
Men HH1 HH2 LN
Children HH1 HH2 LN

Table 3

Summary of Standard Coding Conventions

Response Variable length


Alphabetic One character Two Three Four

Other X 6 characters
96 characters
996 characters
9996
No/None Y na na na na
Inconsistent na 7 97 997 9997
Doesn’t know Z 8 98 998 9998
Missing ? 9 99 999 9999
Not applicable Blank Blank Blank Blank Blank

Since the digits 6 through 9 are reserved for unique purposes, any survey that necessitates

a response of more than six categories should use two-digit answer categories with leading

zeroes (for example, 01, 02, 03, 04, 05, 06, 07, 96, 97, 98 and 99).

In the case of numeric variables, a blank is recorded for a missing value, whereas the

question mark (?) indicates a possible missing value. All instances of the answers "Other"

and "Doesn't Know" are always assigned the values X and Z.

Tools in CS Pro that make it easier to work with century-month codes

Page |7
[Type here]

lcmc = setlb (month, year, minimum); A month, year, and a minimum CMC are sent as

inputs to the function. The CMC is computed

and returned if the year and the month are both

valid. Minimum is returned if the year is invalid.

The CMC for January of the year is given back if

the month is invalid.

ucmc = setub (month, year, maximum); The month, year, and maximum CMC are the

function's inputs. The CMC is computed and

returned if the year and the month are both

valid. Maximum is returned if the year is invalid.

The CMC for December of the year is returned if

the month is invalid

t = adjlba (lcmc, ucmc, di, di, age); The lowest CMC (lcmc), maximum CMC (ucmc),

CMC for the interview date (di), and age are the
if t < 0 then
arguments for the function. The function uses
e = errmsg (2171);
age and di to increase lcmc. If the resultant date
else
is higher than or equal to lcmc and less than or
lcmc = t; equal to ucmc, the function returns the resultant

endif; date. The method returns lcmc if the resultant

date is smaller than lcmc. The function returns -

1 if the output date exceeds ucmc.

t = adjuba (lcmc, ucmc, di, di, age); The lowest CMC (lcmc), maximum CMC (ucmc),

CMC for the interview date (di), and age are the
if t < 0 then
arguments for the function. The function uses
e = errmsg (2171);
age and di to reduce ucmc. If the resultant date is
else
higher than or equal to lcmc and less than or
ucmc = t; equal to ucmc, the function returns the resultant

date. The function returns ucmc if the resultant


endif; date exceeds ucmc. The function returns -1 if the

output date is smaller than lcmc.

Table 5

Sample Cluster Summary Sheet

Cluster #: 101

Total Household Questionnaires ____________

Total Households Completed

____________

Total Women's Questionnaires ____________

MAIN DATA ENTRY

Table 6

Page |9
[Type here]

Office Checks

1. First, make sure that the clusters of Household Questionnaire responses are

ordered by household number, from lowest to highest.

2. In the first section of each home questionnaire, label the Household Information

Panel with the appropriate codes for the respondent's location and the kind of

interview being conducted.

3. Third, the Household Listing Form should be used to confirm the interview

eligibility of every member of the household. Women's modules are open to

anyone who are female and between the ages of 15 and 49. All male participants

must be between the ages of 15 and 49 to take part in the male modules. In order

to participate in the Child Under - 5 courses, one must be younger than five years

old. Each qualified individual's line number should have been coded with the

appropriate eligibility indicator. HL7, HL7A, and HL7B stand for the "Household

Listing Form," an industry standard for exchanging health care information

between providers and patients. Any blanks in the Household Listing's eligibility

areas for new household members should be filled in with the number 00. Issues

should be fixed in accordance with the guidelines outlined in messages 0101–

0133.

4. There can be no discrepancy between the number of eligible women, men, and

children listed on the Household Listing Form and the number of eligible women,

men, and children listed on the Household Information Panel.

5. Number of women's questionnaires, men's questionnaires, and children's

questionnaires in a home must add up to the same total as the number of women,
men, and children who are eligible, as documented on the Household Information

Panel.

6. Result code '1' ('Complete') indicates that all members of the household (women,

men, and children) were interviewed. The number of complete interviews for each

member of the household must equal the number of interviews reported on the

Household Information Panel.

7. 7. Arrange the Questionnaires for Women first, in ascending order of the women's

line number (WM4), then the Questionnaires for Men (in ascending order of the

men's line number (MWM4)), and finally the Questionnaires for Young Children

(in ascending order of the child's line number (CYQ4)) (UF4).

8. Household Listing Form totals must match the Household Information Panel's

total count of women, men, and children that are eligible for inclusion in the

household.

Equal numbers of female, male, and kid household members must be present to fill out a

full set of surveys for each sex.

Complete interviews with women, men, and children in the family must match the

number of responses to the corresponding questions on the Household Factors

Relevant.

Verify that all Questionnaires for Women are grouped together and arranged in

increasing order of the female's line number (WM4), then all Questionnaires for

Men are arranged in order of increasing of the men's line number (MWM4), and

P a g e | 11
[Type here]

finally all Questionnaires for Children Under the Age of Five are arranged in

ascending order of the child line number (UF4).

THE MENU FOR ENTERING DATA SIMPLIFIES THE PROCESS. TOOLS FOR

ENTERING THE CSPRO APPLICATION. THE DATA ENTRY MENU IS

GENERATED BY CH (YOU DO NOT NEED TO MODIFY THIS APPLICATION).

THE BELOW DIAGRAM DEPICTS THE DATA ENTRY MENU. MENU FOR INPUTTING

INFORMATION
DIRECTORY ORDER IN A COMPUTERIZED DATABASE

THE C:MICS5CSPRO DIRECTORY, OR ONE OF ITS SUBDIRECTORIES, CONTAINS EVERY

APPLICATION AND FILE NEEDED FOR DATA INPUT. DATA, ENTRY, AND VERI ARE THE NAMES OF

THE CORRESPONDING SUBFOLDERS. IMPORTANT FILES THAT HAVE BEEN ENTERED INTO THE

COMPUTER MAY BE FOUND IN THE DATA DIRECTORY. IN ADDITION TO ENTRENCH, A SINGLE

BINARY DATA INPUT APPLICATION, ENTRY MENU.ENC IS ALSO INCLUDED IN THE DIRECTORY, AND

IT IS USED TO CREATE MENUS FOR ENTRENCH. EVERY PIECE OF COMPUTER-ENTERED

VERIFICATION INFORMATION MAY BE FOUND IN THIS DIRECTORY.

COMPUTER FILE ORGANIZATION FOR A DATA INPUT CLERK:

MICS5 \S CSPRO \S DATA

PRINCIPAL ENTRY IN FILE SYSTEM

VERI IS A DATABASE OF DATA INPUT APPLICATIONS.

DATA FOR VERIFICATION

THE SUPERVISOR’S MENU

The MICS5 information system is operated by the supervisor via the supervisor's menu.

The CSPro software's main menu is where it's activated. Initial processing from the menu

P a g e | 13
[Type here]

requires entering the cluster number. The picture below depicts the supervisor's menu

once the cluster number has been input, and is followed by an explanation of each choice.

. MODIFY GPS INFORMATION (CHOICE J)

Using this method, the information processing supervisor can launch the program and

modify the GPS coordinates.

CHOOSE A NEW GROUPING OPTION

By selecting this option, the computational supervisor can change the cluster number and

begin processing a new cluster.


Viewing a cluster's tracking data is the fifth available option.

If selected, information about each cluster will be displayed in accordance with its entry in

the cluster tracking form.

OPTIONAL FIELD CHECK TABLES

When this checkbox is selected, the EXPORT COMMAND will be used to export all of the

files in the RAW SUBFOLDER. A total of nine ASCII text files and corresponding SPSS

description files are generated by the program. Once these files have been imported into

SPSS, you may use the corresponding SPSS syntaxes (MICS5 - TABLE F1) to create field

check tables. SPS

Choose option R to replicate the operator's key's data file.

This choice copies the cluster data file from the manufacturer's key to the supervisor's

computer in the event that a network cannot be created.

Copy the verification file with the operator's key by selecting Option S.

This choice will transfer the cluster verification data file from of the operator's key to the

superintendent's computer in the event that a connection cannot be created.

P a g e | 15
[Type here]

Organizational structure of the supervisor's computer:

Directory for MICS5 CSPRO backups Data dictionary index ENTRY Data-entry

applications index EXPORT The last step involves exporting both the application directory

and the data directory. GPS Software for entering GPS coordinates Data in their original

format When updating the application directory, MATMORT Maternal Mortality

Imputation

IT IS IMPORTANT THAT THE DATA HAVE A VALID UNDERLYING STRUCTURE.

Although the data input method has its limitations and cannot guarantee complete

accuracy, it does ensure that the majority of fundamental coherence is maintained. Because

of this, a structural validation program must be executed once the primary information

entry is finished. The structure checking program does more than just verify that the

number of questions in the data file matches the number of questionnaires received from

the field; it also puts each questionnaire through a battery of tests to ensure its integrity.

PROOF READING THE STRUCTURE


MICS5 Data Structure Check

Cluster: 3

Households | Women | Men | Children |

Total Comp Income | Eligible Interviewed | Eligible Interviewed | Eligible Interviewed

| HH12 FOUND HH13 FOUND|HH13A FOUND HH13B FOUND |HH14 FOUND

HH15 FOUND

2 1 1 |5 5 4 4|6 6 5 5 |4 4 3 3

In addition, the checkback program compiles a complete directory of all cluster homes. If

there was a household interview, the total number of women, men, and eligible children

who filled out a questionnaire is shown together with the household number and result

code (that is, the household result code is equal to 1). What follows is an illustration of a

typical domestic unit's output.

Family Member 1 for MICS5 Data Structure Verification

The Outcome Was 1

Eligible Interviews Conducted with Female | Male | Junior Participants

Found: HH12, HH13, HH13A, and HH13B, as well as HH14 and HH15

4433|5544|2211

P a g e | 17
[Type here]

FILE ANALYSIS CREATION

The basic data processing output file is structured in a way that makes it easy to enter data.

This structure is not best for analyzing the gathered data, thus recoding variables is the

first step after transferring the data to SPSS. Making an analysis file is the term for this

process. The methods for creating MICS5 analysis files are described in great depth in this

section. Analysis files created using these methods can be shared with researchers for use in

model tabulation.

Women who have heard of AIDS 10


Total number of women 20
Percentage of women who have heard of 10 / 20 x 100 = 50

AIDS of the variable AIDS


Mean (10 x 100 + 10 x 0)/20 = 10 x 100/20 = 10/20 x 100 = 50
Table 7

Checklists

Before fieldwork:

 Get your hands on some new computing devices.

 Create a room or area that serves only as a data processing hub.

 Hire a data processing supervisor and extra helpers.

 Put in place a method for classifying different types of processing work.

 The pre-test survey's findings should inform any necessary adjustments to the

programs.

 Analyze and refine your survey based on early results.

 Pre-test and post-test questionnaire data should be used to evaluate the success of

the program.

 Outsource the bulk of your data entry needs while you're out in the field.

 The primary data file should be checked for proper organization.

 Task someone with checking the data entry.

 Make that the two sets of data, the main and the verification, are same.

 The raw data file must be saved in a safe location.

 Field checks on tables, please.

 Refine the supplementary materials.

 a second copy of the finalized data set.

P a g e | 19
[Type here]

 Following data collection: Assume conflicting and insufficient information; export

to SPSS.

 Adjust the settings, please.

 Design a financial well-being index, complete with GPS coordinates and illustrative

computations and weight increases.

 Conduct the tabulation procedures.

 Create a plan for data exchange and storage procedures.

 All applications, documentation, and analytical data should be forwarded to

UNICEF..
PROCESSING THE DATA 21

Table 8

Sample Cluster Tracking Form

Number of questionnaires Data entry Date Date

of of

Cluste Date Date raw Date final

r Date Househol Operat Operat check verificatio data of

numb receive ds Househol Women Men’ Children or or Creatio complet n back editin back

er d selected d ’s s ’s name number n date e complete -up g -up

M:

V:

M:

V:

M:

V:

M:
PROCESSING THE DATA 22

V:

M:

V:

M:

V:

M:

V:

M:

V:

M:

V:

M:

V:

References

Earl Babbie is a. conducting social research in practice. tenth revision. 2003; Wadsworth Publishing.
PROCESSING THE DATA III.1.23

Print. 174 Dana John Creswell and Lynn Driscoll. Qualitative, quantitative, and mixed method approaches to research design. 3rd ed. 2008

Sage publications

Rubin, Herbert, and Irene Rubin. Print. The art of hearing data through qualitative interviewing. second edition Sage Publications,

Thousand Oaks, CA, 2004.

Fink, Arlene. Print. A Step-by-Step Guide for Conducting Surveys 2008, Thousand Oaks, California: Sage Publications, 4th ed.

Sanger, Jack. Print. Fully Observer? A Field Guide to Observation for Research. 1996, Routledge, New York.

The National Commission for the Protection of Human Subjects in Biomedical and Behavioural Research should be printed. 18 April 1979:

The Belmont Report. Web

You might also like