Scribd
Upload
926 views

Bias in Data Collection

This document discusses different types of bias that can occur in data collection, including sampling bias, non-response bias, measurement bias, and response bias. Sampling bias occurs when the sampling frame does not accurately reflect the population being studied. Measurement bias results from issues with how the data is collected, such as poorly calibrated instruments or leading questions. Non-response and response biases involve participants not providing accurate responses due to lack of participation or deliberately misleading answers. Addressing these biases is important for obtaining representative and reliable data from surveys and studies.

Uploaded by

api-204699162
Copyright
© © All Rights Reserved
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
926 views

Bias in Data Collection

This document discusses different types of bias that can occur in data collection, including sampling bias, non-response bias, measurement bias, and response bias. Sampling bias occurs when the sampling frame does not accurately reflect the population being studied. Measurement bias results from issues with how the data is collected, such as poorly calibrated instruments or leading questions. Non-response and response biases involve participants not providing accurate responses due to lack of participation or deliberately misleading answers. Addressing these biases is important for obtaining representative and reliable data from surveys and studies.

Uploaded by

api-204699162
Copyright
© © All Rights Reserved
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 14

Bias in

Data Collection

BIAS IN DATA COLLECTION


The results of using a sample to make
predictions or inferences about a
population are only accurate if the sample
is representative of the population.
The methods for choosing the sample and
collecting (measuring) the data must be
free from bias

BIAS IN DATA COLLECTION


STATISTICAL BIAS is any factor that
favours certain outcomes or responses which
systematically skews the survey results
A BIASED SAMPLE is a statistical sample
in which members of the population are not
equally likely to be chosen

SAMPLING BIAS
Sampling bias occurs when the sampling
frame does not reflect the characteristics of
the population
Sampling Frame the group of individuals
that actually have a chance of being selected
from the population

SAMPLING BIAS: Example


An ad agency in a developing country wants to
know what proportion of households have at
least one computer. The survey was conducted
by randomly selecting households from the
telephone directory.
Why is this a sampling bias?

SAMPLING BIAS: Example


An ad agency in a developing country wants
to know what proportion of households have
at least one computer. The survey was
conducted by randomly selecting households
from the telephone directory.
Why is this a sampling bias?

SAMPLING BIAS: Example


Why is this a sampling bias?
It is rare for households in developing
countries to own a telephone. Those who
have a phone may have the money for a
computer. Selecting households from the
telephone company directory is creating a
sample the sample that is biased and
overstates the proportion of the population
owning at least one computer.

SAMPLING BIAS: Example


There could be a significant number of
households without telephones
May over-estimate the proportion of
households with computers
A better way would be to visit randomly
selected homes, however, it may be more
expensive

NON-RESPONSE BIAS
Occurs when particular groups are underrepresented in a survey because they choose
not to participate

A voluntary response sampling technique can


lead to on-Response Bias
People least interested in the issue will tend
not to respond to a survey

MEASUREMENT BIAS
Occurs when the data collection method
consistently over- or under-estimates the
characteristic of the population
A consistent measurement error which skews
the results of the survey. Examples:
(1) When measurement instruments are poorly
calibrated, affecting the result
(2) When questions in a survey lead to a
particular answer or response

MEASUREMENT BIAS: Instrumental


Results from poorly measuring the outcome
you are measuring. Examples:
The survey interviewers asking about deaths
were poorly trained and included deaths
occurred beyond the time period of interest.
One survey team's portable scale to measure
participants weights malfunctioned and was
not checked, adding 0.5 lbs. to actual weight.

MEASUREMENT BIAS: Leading Question


A question that suggests an answer which
otherwise may not have been chosen by the
respondent without prompting
For example, multiple choice questions with
suggested answers:
What is your favourite video game?
a) Mario Striker
b) Tetris
c) ESPN NHL 2K10
d) other

MEASUREMENT BIAS: Loaded Question


When a question contains wording or
information which influences the way the
respondent will answer
For example, Not having MP3 players in
class will reduce the number of distractions
to students in the class. Are you in favour of
raising the achievement levels of students by
banning MP3 players from the school?

RESPONSE BIAS
When participants in a survey deliberately
give false or misleading answers
Respondents may want to influence the
outcome of the survey results or they may be
afraid to answer the question truthfully out of
embarrassment or sensitivity

For Example, Teachers asks Who has ever


cheated on a test?

You might also like