Research Glossary
Think of this as an encyclopedia of personnel research. I will
be adding to it continually throughout the semester. If you have
suggestions about items to add or changes in the definitions,
please email me at steve_borgatti@msn.com
or fill out the comments form -- I
would really appreciate it!
- Applied Research
- 1. The kind of research usually performed by consultants
or HR professionals. Typically motivated by the need to
solve a specific problem in a particular organization. 2.
Contrasts with basic research.
- Basic Research
- 1. The kind of research usually done by academics.
Typically tries to uncover universal relationships among
variables. This kind of research generally has
implications for solving particular problems in specific
organizations, but is primarily motivated by a generic
desire to understand how things work. 2. Contrasts with applied research.
- Cases
- Objects or entities whose behavior or characteristics we
study. Usually, the cases are persons. But they can also
be groups, departments, organizations, etc. They can also
be more esoteric things like events (e.g., meetings),
utterances, pairs of people, etc. In the context of
sampling, cases are also called elements.
- Causality
- While the goal of research is to understand what causes
what, this is a very difficult goal to achieve. Strictly
speaking, it is impossible. In fact, the notion of
causality is just a theory itself. However, on a
day-to-day basis, we assume that causality does exist and
that we can discover it through a combination of
inductive and deductive work. In general, laboratory
experiments are the only way to ascertain causality.
- Cluster Sampling
- A multi-stage sampling scheme in which the population is
first divided into clusters, then a sample of these
clusters is chosen via simple random sampling, and then a
simple random sample of population elements is selected
within the chosen clusters. This differs from stratified
sampling in that in stratified sampling, all strata are
sampled, whereas in cluster sampling we take a sample of
clusters, not all clusters. Cluster sampling is used when
it is difficult to construct a sampling frame for the
entire population, or when it is too costly to visit
randomly chosen population elements.
- Construct (noun)
- A variable in a theory. Sometimes carries the connotation
of something that cannot be observed directly, or which
we suppose to exist but has not been measured yet.
Similar in this sense to a latent variable. Intelligence
is a construct that is used to explain competence.
- Cross-Sectional Study
- A cross-sectional study is where we collect data only
once from each unit of analysis. For example, if we want
to examine the effects of age on attitude towards
abortion, we collect attitude data from people of all
ages, then see if there is a correlation between age and
attitude. This is the opposite of a longitudinal study, where
you take a set of young people, then measuring their
attitude towards abortion every few years as they get
older.
- Data
- The outcome of measurement.
The set of values or codes that record what was observed,
such as the blood pressure of 100 people.
- Descriptive Study
- A descriptive study is similar to an exploratory study in
that we do not attempt to test hypotheses. Often, they
are used in settings where a theory of how variables are
related is already in place, but specific values for each
of the variables are needed for specific cases in order
to take some action. For example, if an organization is
considering adopting a new benefits package that costs
more but has new features that might be attractive, the
organization needs to know what the needs of the
employees are to determine whether the package makes
sense for them. For example, if the main feature of the
more expensive package is a domestic partner program
(that's where homosexual partners of employees are
entitled to health insurance just like heterosexual
spouses), it makes sense to find out how many gay &
lesbian employees the firm has.
- Elements of a Population
- In the context of sampling, elements are cases -- units
of observation. They are the things being sampled. In
general, elements are persons.
- Exploratory Phase
- The exploratory phase of a study is where you try to
figure out (usually qualitatively) what is going on.
There are three basic objectives: (1) learn the lingo of
your respondents; (2) learn the background context within
which everything happens; and (3) develop a set of
testable theories about what is going on
- Exploratory Study
- Also known as "inductive" or
"theory-building". In this kind of study, we
don't begin with a theory. Instead, we collect data that,
after analysis, we will use to develop a theory. After we
develop the theory, we might then design a study to test
the theory.
- Falsifiability
- In the context of theory construction, is the property of
a theory to possibily be shown false. There are several
ways to be non-falsifiable. One is to be circular
(tautological). For example, to explain why people do
dumb things, you could theorize that it is because they
are dumb. If you then define dumb as a person who does
dumb things, the theory is circular. Another way is to
appeal to things that can't be measured. For example, to
explain why people go to the movies, we could theorize
that they want to. But there is no way to measure whether
they want to that would avoid circularity.
- Field Experiment
- A field experiment is a study in which you make changes
in the independent variable to see how it affects the
dependent variable, but otherwise you leave everything in
its natural state. For example, an ethologist (someone
who studies animal behavior), might put sugar water at
different distances from a bee nest in order to observe
how the difference in distance affects the dance that the
bees do on returning home to communicate where food is.
- Field Study
- A field study is a study in which the researcher goes to
a research site and observes and asks questions, but does
not change anything. It's like a naturalist observing
wildlife without doing anything like setting out food to
attract animals, or making obstacles to see how the
animals react.
- Haphazard Sampling
- A non-probability sampling scheme in which population
elements are chosen based on convenience (e.g., choosing
your friends to make up a sample of college students).
- Hypothesis
- 1. A postulated relationship between a pair of variables.
The reason for expecting the variables to be related
should come from a theory. 2. Any
theory-based prediction about some measurable data.
- Interval level of
measurement
- 1. A level of measurement in which the ratios of measured
values are not meaningful, which is to say that they do
not correspond to similar relationships among the objects
measured. The classic example of interval measurement is
the measurement of temperature using Fahrenheit and
Centigrade scales. If it is 80ºF in Tulsa and 40ºF in
Juneau, you cannot say it is twice as hot in Tulsa. Here
is one clue that this is not meaningful. Suppose we
measured the temperature in Centigrade instead of
Fahrenheit. The formula for converting Fahrenheit to
Centigrade is C=(F-32)*5/9. So in Tulsa it is 27 C and in
Juneau it is 4 C. Now it looks like Tulsa is 4 times as
hot as Juneau! Yet Fahrenheit and Centigrade are
perfectly equivalent and equally valid measuring scales.
So you know ratios are not meaningful in interval
measurement. 2. The ratios of differences (intervals)
among measured values is meaningful. For example, suppose
it is 70 F in L.A. and 50 F in San Francisco. The
difference in temperature between Tulsa and Juneau (40 F)
is twice as much as the difference in temperature between
L.A. and San Francisco (20 F). This statement is still
true if we measure the temperatures in Centigrade, so
interval measurements preserve ratios of differences
in measured values.
- Intervening Variable
- An intervening or intermediary variable is one that is
affected by the independent variable and in turn affects
the dependent variable.
- Judgment Sampling
- A non-probability sampling scheme in which you make use
of special expertise to select elements for the study.
Typically, this is used to obtain a balance of viewpoints
or to select knowledgable respondents.
- Laboratory Experiment
- A lab experiment is a study in which you make changes in
the independent variable, and then control all the other
variables so that only the variable of interest could
possibly affect the outcome. For example, if you are
interested in the effects of seeing an inspirational film
on taking a math test, you can recruit some experimental
subjects to come to your theatre, then randomly assign
half to see the film, and the other half to see some
other film, then give them the test right after the
films.
- Level of measurement
- At the simplest, level of measurement refers to what
kinds of arithmetic relationships among the numeric
values of the data actually reflect some kind of similar
relationship among the objects themselves. Although there
are many different kinds of measurement, in this course
we will pretend that there are just 3 levels of
measurement: ordinal, interval, and
ratio.
- Longitudinal Study
- A longitudinal study is where we follow the units of
analysis (say, employees) over time, and measure key
variables at different points in time. For example, we
might measure morale before and after a promotion.
- Measurement
- The generation of data. A process of assigning numbers
(or codes) to things such that certain specifiable
relationships among the things are reflected in certain
relationships among the numbers. For example, when we
measure the mass of objects, we assign a number to each
object, known as its weight. If the number assigned to
object A is 10 and to object B is 20, we can say that
object B has twice as much mass as object A. This
preservation of ratios works for the way we measure mass,
but it doesn't work for the way we (usually) measure
temperature. Exactly which kinds of relationships between
the numbers is actually reflective of relationships among
the objects is what defines the level of the
measurement.
- Moderator Variable
- A moderator variable is one that modifies the
relationship between two other variables. If variable X
modifies the relationship between variables Y and Z, then
there is an interaction between X and Y. In a regression,
the interaction between a pair a variables is tested by
including the product of the two variables as an
additional independent variable.
- Non-Probability Sampling
- Any sampling scheme in which the probability of a
population element being chosen is unknown. There are
four basic kinds: haphazard, quota, judgement, and
snowball.
- Population
- In the context of sampling, population refers to the
universe of all possible cases. If you are studying the
members of IBM, it is the set of all members of IBM. Can
be used in contrast to sample.
- Parameter
- 1. A summary value calculated from a population (as
opposed to a sample). Contrasts with statistic.
- Probability Sampling
- Any sampling scheme in which the probability of choosing
each individual is the same (or at least known, so it can
be readjusted mathematically to be equal). Also called
random sampling. Probability samples are more costly to
obtain, but are more accurate, and they allow the
researcher to calculate the amount of error she can
expect. There are three major kinds of probability
sampling: simple random sampling (SRS), stratified
sampling, and cluster sampling.
- Questionnaire
- A formal, written, set of closed-ended and open-ended
questions that are asked of every respondent in the
study. The questions may be self-administered, or
interviewer-administered. A source of data.
- Quota Sampling
- A sampling scheme similar to stratified sampling in which
you first divide the population into classes (such as
males and females) and then obtain a haphazard sample
within each class.
- Sample
- A subset of population elements. In some usages,
contrasts with population.
- Sampling
- The practice of choosing a subset of population elements
to study instead of the entire population. In general, we
sample because (a) it's cheaper; (b) in some cases the
population is theoretically infinite. There are two basic
kinds of sampling: probability and non-probability.
- Sampling Error
- A difference, due to sampling, between a population
parameter and the corresponding sample statistic. For
example, the average age of a population might be 25
years, but a given sample might yield an average of 26
because, by chance, more old people were selected than
the population proportion.
- Sampling Frame
- The sampling frame is a specific list of names (or other
identifying codes) of the cases to be sampled. Usually,
this is supposed to be the same as the population. For
example, when you study IBM, you start by obtaining a
list of all IBM employees. This is the sampling frame. If
your list is not complete (e.g., it omits top
management), your results may not be valid in the sense
of generalizing to all of IBM.
- Sampling Ratio
- The sampling ratio is the size of the sample divided by
the size of the population.
- Simple Random Sampling (SRS)
- A sample in which the population is first divided into
strata (classes of elements). Within each stratum, each
element has an equal chance of being chosen for the
sample.
- Snowball Sampling
- A non-probability sampling scheme in which you begin by
sampling one person, then ask that person for the names
of other people you might interview, then interview them
and obtain a list of people from them, and so on.
- Statistic
- A statistic generally refers to a summary value
calculated from a sample. For example, we might compute
the average age of all respondents. The term contrasts
with population parameter.
- Stratified Sampling
- A sample in which each element in the population has an
equal chance of being chosen for the sample.
- Theory
- A general explanation of how something works. A theory
says what is related to what and why. A theory is, in
part, a collection of related hypotheses. However, a
theory also contains a sense of process and mechanism --
a sense of understanding of why and how the variables are
related the way they are. Desirable characteristics of a
theory include: falsifiability, parsimony, truth,
fertility, generality, surprise, and a sense of process
or mechanism.
- Theoretical Framework
- A theoretical framework is a theoretical perspective. It
can be simply a theory, but it can also be more general
-- a basic approach to understanding something.
Typically, a theoretical framework defines the kinds of
variables that you will want to look at.
- Time Allocation Studies
- A technique for determining how much time people spend
doing different activities. Basically what you do is
arrive at specific locations at random intervals and
record what everyone is doing. Or, you don't arrive at a
particular location but instead go find a specific
individual and record what that individual is doing. For
example, your theory might require you to know how much
time the manager spends on a set of key activities, such
as writing reports, talking in meetings, or socializing.
- Variables
- 1. Characteristics, attributes, or qualities of cases. For example, if the cases are
persons, the variables could be sex, age, height, weight,
feeling of empowerment, math ability, etc. 2. Anything we
measure. 3. Constructs in a theory.
-