CHAPTER 3
3.1 How do you express constructs in measurable terms?
How are you going to gather the data you need? The quality of the data that is gathered
is defined by how that data is gathered. Important:
It is important to use a proper definition of whatever construct you are measuring.
This ensures that everyone who reads your research report correctly understands
what it is that you set out to measure.
It is important, particularly for abstract, complex constructs (satisfaction,
happiness, customer-friendliness) to look for concrete constructs you can use as
an indicator for measurement.
This process of translating abstract construct into concrete, measurable terms =
operationalization.
Operationalization translates the construct as intended (a construct like intelligence)
into a construct as measured (IQ score on an intelligence test). Once a construct is
translated into something that can be measured: it is referred to as a variable (= a
measurable construct which can be expressed in different scores or values).
3 phases of the operationalization process:
1. Defining
2. Establishing (sub)dimensions
3. Finding indicators
Defining: describing what, exactly, you mean by the construct. TIP: use specialist
literature to define constructs.
Establishing (sub)dimensions: a construct such as communication has different
dimensions (=aspects). Example: verbal and non-verbal. Some of these dimensions can
even be divided into further subdimensions.
Finding indicators: for each dimension, you need to find and apply indicators to let
you measure the concept.
Operationalizing a complex, abstract concept into a number of points for observation is a
sensible approach.
When asking questions in an interview, you need at least a good introduction, a good
opening question, and a topic list. Your topic list in particular relies on the inclusion of
both physical and psychological consequences, including their indicators, but also on an
overview of possible stressors. Even the most qualitative of studies, therefore, requires
due preparation.
3.2 Which data collection methods are suitable?
Research makes use of the following data collection
methods:
, Using existing materials;
Asking questions;
Observing.
Using existing materials: is the preferred choice where data collection is concerned. It
bothers no one, and is therefore not affected by reluctance. Existing data is becoming an
increasingly important source of research data. Example: social media message or
websites you click on or data in a cash desk register. Benefits:
Not having to trouble any respondents
Limits costs
Negates the effort involved with reminders or non-response
Existing materials: articles, tapes, datasets, pictures.
Asking questions: a great deal of information is obtained by asking questions. However,
you should not always rely on a person’s answer about themselves. It may be better to
ask others, instead. This is known as using informants. Ex: studies about managers,
asks the employees under them. Studies about children, asks their parents.
Observing: a study that focusses on behavior should, ideally, make use of observation.
But observation is difficult and often undesired. People are unlikely to appreciate being
observed, and will probably behave differently than normal.
3.3 Is your data collection reliable and valid?
The quality of the data you collect is determined by its:
Validity
Reliability
Relationship between reliability and validity
Validity:
Construct validity
Ecological validity
Construct validity: is the extent to which you measure what you intend to measure.
Term mostly used in quantitative research. If your study does not measure what it is you
are intending to measure, then you are dealing with poor construct validity.
Social desirability: is a response acting in accordance with their perception of the norm
in terms of their answers and behavior. It is a threat to good validity. Example: children
who are abused by their parents are likely to stay loyal to them.
Ecological validity: Stay close to reality (trustworthiness in regard to daily practice).
Term mostly used in qualitative research (higher). Qualitative research generally does not
refer to validity, instead preferring the concept of. trustworthiness or plausibility.
In order to improve the validity of the results, qualitative researchers like to make use of
the principle of triangulation = multiple measures from different perspectives
(increases validity) / measured with different methods If you observe AND interview
different people, then your trustworthiness improves (less subjectivity).
Reliability (stability): is the extent to which a measurement is independent (remains
unaffected) of chance > stable across time and situations. So, reliability is about
consistency. Is your research reliable, then you should get the same results when
somebody else conducts the study or when you do the study again. Has to do with:
Measurement instrument used:
o Low reliability when: measurements instrument used (test, questionnaire
what can interpreted differently, observation)