Introduction: that we must choose designs for picking


The phrase
‘longitudinal’ is used to symbolize a set of studies that are conducted over a
period of time. Often, as we have seen, the word ‘developmental’ is employed in
connection with longitudinal studies that deal specifically with aspects of
human growth (30)

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now

studies (defined in general as studies in which the response of each individual
is noticed on two or more occasions) represent one of the principle research
strategies employed in medical and social science research (Goldstein 1979;
Nesselroade and Baltes 1979). Longitudinal designs are uniquely convenient to
the study of individual change over time, involving the effects of development,
aging, and other factors that affect change. Despite the importance of the
longitudinal study, however, satisfactory methods for the analysis of serial measurements
are not readily available. The statistical literature on the analysis of serial
measurements is based on the paradigm of multivariate regression, and standard
statistical software packages have the same orientation. Yet longitudinal
studies typically have

 Unbalanced designs, missing data, attrition,
time-varying co- varieties, and other characteristics that make standard
multivariate procedures inapplicable. (21)

studies have a long history in medical and social science

Research. They
offer a natural access to the study of development and

Aging that
allows the separation of age and group effects. They can also be

Used to
produce precise estimates of treatment contrasts not subject to

variability. Yet they are often more difficult and necessitate

Greater spending
per surveillance than cross-sectional studies. (21)

The moment
that we have decided to running a longitudinal investigation, we must

Mark exactly
how, when, where, and on whom the measurements will

Be taken.
Longitudinal designs principally have subjects crossed with

This means that we must choose designs for picking out subjects

And also the
occasions at which they will be measured. (21)

The design
of a longitudinal study has two aspects: a design for selecting

Subjects and
a design for occasions.

longitudinal research is characterized by:

The obvious
integration of three elements: (a) a well-expelled theoretical model of

observed using (b) a temporal design that bears a clear and itemized
view of

The process,
with the resulting data analyzed by means of (c) a statistical sample

Is an
operationalization of the theoretical model? Two general varieties of

Models are
considered: models in which the time-related change of primary interest is

and those in which it is characterized by movement between unattached states.

One of the
scientific benefits of directing longitudinal study is the ability to observe
temporal order of a key exposure and outcome events ,particularly we can locate
what ever changes in a covariate precedes changes in the outcome of the
interest (25).

A possible hardness
in longitudinal study is that the measurement of defendant may be missing at
one or more time points (32)

The longer a
study keeps the higher will be the value of the resulting longitudinal

Data sets.
The clear reason is the permanent collection of information concerning

developments. (8)

designs have two primary motivations:

1. To raise
the regulation of treatment contrasts by eliminating interindividual

This is obtained by watching each subject under the

treatment (or exposure) situations to be compared. Such designs

Are called
repeated measures designs, and involve the cross-over design

As a special
case. Repeated measures designs use each subject as his or,

Her own

2. To test
the individual’s coverting response over time: Longitudinal

Designs have
natural entreaty for the study of changes correlated with

or aging. They have value for describing both temporal

Changes and
their depending on individual characteristics.(22)

new research
has focused on the development of statistical methods that not only take into
account the inter- correlation of serial measurements but also harmonize the
complexities of typical longitudinal data sets and declare the specification of
mean-value functions determined by subject matter considerations rather than by
constraints introduced by the statistical methodology. This work has been based
on the concept that the analysis of serial measurements should be inspected as
a univariate regression analysis of responses with correlated errors. (21)

studies are distinguished by frequent observation of individual respondents.
(The terms respondent and study participant are used because of our primary assurance
on human investigation. The methods described, however, apply quite generally
to longitudinal research, and the term experimental unit has the same technical
meaning.) The metameter for the occasions of measurement may be age, time on
study, or some other natural scale; alternatively, the relations of measurement
may match to levels of an experimental or observational variable, possibly with
no natural order in. (21)

The excellence
in longitudinal research requires care in both the

Design and attitude
of studies. The general goal in conducting the study is

To perform
the study as designed. Three important objectives in performing

studies are the following:

1. Bring
down and quantify instrument and observer variability.

2. Guarantee
rigorous recording, coding, and transcription of data.

3. Bring
down nonparticipation and obtain complete data at regular visits (23)

studies give more precise rating of temporal changes

Or treatment
effects than cross-sectional studies of the same size.

They attain
this gain in precision by completely removing interindividual variability from
the comparisons of interest.














A study
design is a project for a data set that will efficiently test study

and evaluate important parameters. Since analyses depend on

The data
actually collected; however, the importance of a study depends on equally

On its
design and its implementation. Although the details of a well-conducted

Study based
on the variables measured and the goals of the study, some

for a well-executed study cut across disciplinary lines.

In any study
involving measurement, sources of bias and measurement

Error must
be controlled. Generally, this requires inclusive training of

and evaluation of the validation and reliability of measurement

Tools, for
example, in the Hypertension Detection and Follow-up Program

A study in
which blood pressure level was a critical measure of treatment

random zero sphygmomanometers were used to avoid bias.

Arising from
digit preference or other subjective factors, Blood pressure

devices were rigorously pretested, and observers were needed

participate in a one week training course and pass a certification

these procedures were intended to minimize interobserver

interinstrument variability.

In a
longitudinal study, there is an additional need to maintain study

over time, Instrument performance must be constantly watched

systematic bias re-estimated. Similarly, investigators should plan

For training
of new observers and recertification of settled observers to

deterioration of measurement procedures. Because this element of a

Study does
not participate directly to study results, it is a natural candidate

For reduced
effort during periods of tight budgets. To avoid this mistake,

designs and study budgets should include sufficient support for this

Work, Investigators should set a regular schedule for checking

And observers
and the results of these checks should be kept as

A part of
the study records. When proper, instrument and observer

should be recorded with each observation, so that these variables

Can be regarded
in the analysis if necessary, In some studies, instrument

And observer
variability are comparable in significance to the effects under

Study and
may bias estimates of effects of interest if not cautiously controlled.

Bias can
also arise in the recording, transmission, and entry of study data.

many studies confirms extensive quality control for data entry

To computer
files, variability in reading participant records such as Spiro grams

Or coronary
angiograms and inconsistency in coding questionnaire

Data often
represent more important sources of variability. These sources

uncertainty can be quantified and controlled only by extra quality

activity in the data collection system, although observer variability

Is a widely known
phenomenon, specialists are often surprised by the

Amplitude of
observer variability in interpreting diagnostic tests such as

duplicate readings should be introduced to quantify

This variability
and blinded reading should be required when there is potential

For bias in
comparative work,

well-designed longitudinal study can also be intimidated by problems

Of missing
data, Data can be missing either because of procedural error

During a
regular visit or because a participant does not appear for a regular

Visit. Both
events are harmful to the study. Most procedures for treating

Missing data
in the analysis suppose that data are missing at random (44),

I.e. the
probability of missing an observation does not depend on the value

of that
observation. When that hypothesis is true, the main outcomes of missing data
are (a) obstacle, because unbalanced data sets are

difficult to analyze, (b) loss of regulation, because missing outcomes

Decrease the
effective size of the study, and (c) problems with modification for

when their values are missing.

longitudinal studies, the proposition that observations are missing at

Random is
frequently not reasonable. Participants who are lost to follow-up are

atypical in terms of mobility, social class, and general health. This is

A special
threat to comparative studies in which different groups have

follow-up processes, this problem is overcomes by introducing

procedures to control follow-up of study participants and promote

at regular visits.(22)

longitudinal studies are studies which include repeated measurement

Of the same
individuals over time. Panel analyses are an example. Experiments, in

Which a
particular „treatment** is given to a portion of the study group

repeated measures of the same subjects, could in principle be seen as of

design as well. However, as long as the before and after design involves

A short time
period only, we excel not to include experiments. Prospective longitu¬

studies might take on various forms according to their variables: If the same

are repeatedly measured using the same variables the term panel analysis

Is usually
applied. If the same individual is repeatedly measured using different

the term prediction studies are not uncommon. Both kinds of longitudinal

mix, since even in prediction studies, conceding marital adjustment

delinquency for instance, some variables remain the same in the subsequent

waves. possibly for this reason, many authors see the terms panel and

study as equivalent

longitudinal studies might also take on different forms with regard

To the time
dimension of sampling. One may start from data collected in the past

By other
researchers and follow them up. One could, of course, also start from the

Present and forward
to the past for given individuals by taking recourse to archival


retrospective or quasi-longitudinal studies the subjects have only one measure¬

ment in time
but data attaching to various points in time , decisions and definitions they
brought to the Situation in past times.

longitudinal approach in this design is thus one depended on the memory

Of the
respondents, various data collection strategies have been associated with this

ranging from more or less unstructured forms of data collection (such as

qualitative analysis of specially triggered written autobiographies)

To highly
structured closed interviewing Samples have accordingly varied be¬

tween same
samples of special social groupings to large samples of the population in


longitudinal studies are started or continued the most significant practical

Problem mostly
to be solved is retrieving the respondents. Depending on the char¬

of the population, the records with which one is starting, and the time

that has passed,
the number of cases which manifest to be lost at the outset is nor¬

mally very high.

There can be
no doubt, nevertheless, that some studies — according to sample and de¬

characteristics — have bigger chances than others to succeed, regardless of the

used to determine respondents. Leading factors affecting tracing failures are

the size,
mobility and scuttle of the sample. Among social groups relatively ho¬

and centrally located populations with high education (such as universi¬

ty students)
the researcher probably fairs best. The chance of locating them might

be high
simply because people tend to stay in these institutions for some time,

contacts with many people there, and later follow certain relatively homo¬

geneous career
patterns. They might, moreover, belong to alumni Organization and

professional organizations, whose directories might be searched. High education

furthermore raises cooperativeness.(40)


To analyze
longitudinal data, one must appoint the probability distribution

for each
subject’s set of responses. We shall at first suppose that the outcome

Variable is
a measurement that has a normal distribution,

The approach
to modeling the predictable outcome depends on the goals

of the
study. In repeated measures studies, differences between occasions in

the expected
outcome for a single subject are refer to changes in

Treatment or
exposure conditions, the analysis starts by testing the equality

of mean
values over occasions and continues with estimation of the

in means between occasions.

When changes
in the predictable outcome over occasions are due to growth

or aging. or
when occasions coincide to different levels of exposure, the

analyst will
want to model the changes over occasions. (22)

missing data in longitudinal research can occur because a participant

Fails to
respond to one or more questions in a questionnaire or interview, or because

participant is not obtained to the research study at one or more opportunity of

Scientists involved in longitudinal research deal with unplanned

missing data
permanently; in fact, it is difficult to visualize a longitudinal study without

at least
some unplanned missing data. For this reason, how to handle missing data

is an
important question meeting anyone who wants  to analyze longitudinal data.

For years
investigators have used ad hoc procedures for dealing with missing

data, such
as eliminating individuals with missing data from analysis (“casewise

or substituting the sample mean for missing observations (“mean substitution)

procedures may be adequate, but they have no basis in statistical theory. There
are two prospect consequences of using ad hoc procedures to deal

unplanned missing-ness. One is a higher-than-necessary loss of statistical

particularly in association with case-wise deletion, which including discarding

data for any
subject whose data are incomplete. The other consequence is bias

in results,
which can occur if the cause of missing-ness is linked to variables of


A much advanced
option is to use modern missing data procedures (Schafer 1997,

& Graham 2002), such as multiple imputation and maximum likelihood,

which are depends
on statistical theory. When the assumptions underlying these procedures

are met,
they restore much statistical power and eliminate bias due to

data; even when the underlying assumptions are not met, modern missing

procedures are an improvement over ad hoc methods (Collins et al. 2001).

This is
particularly so if variables that are highly correlated with those subject

to missing-ness
are involved in the analysis. Commonly in longitudinal studies,

Past or
later measures of a variable may render this role well. Collins et al.

(2001) and
Graham (2003) clarified why and how to implement this the better approach, and pretend
how such a sample can be tremendously

for making the most of modern missing data procedures in longitudinal


The longer a
study lasts the higher will be the value of the resulting longitudinal large-scale

data sets.
The clear reason is the permanent accumulation of information concerning

developments. Moreover, in educational research there is a obvious need

longitudinal large-scale assessments because competence development and

paths can be
watched out for longer time spans. Long-lasting studies are surely

even if a lot of time and resources are invested in keeping a panel stable and

diminishing biasing effects. Therefore, the design of longitudinal large-scale

should not
remain fixed over the whole course of the study but should be adapted from

time to
time. In this direction two main levers exist. The time spans between

can be
widened basically, possibly with panel care measures in between. Alternatively,

a panel can
be discontinued in order to invest the resources in a restart of a new panel

with a fresh
and unbiased sample. This is especially valuable when, for example, fundamental

changes have
occurred in society or in settings that are of study interest, and

which would
otherwise be insufficiently covered.(8)

in spite of
the fact that missing data has the chance to cause serious bias, it is still
possible to carry out a adequate and rational analysis. In order to do so,
however, interest must be given at each stage of the research: design, data
collection, statistical analysis and reporting. in order to address the issue
of missing data in patient recording outcome, including key definitions,
prevention practices, and analytical approaches, including sensitivity
analyses. We will not discuss questionnaire development, and make the tacit supposition
that the patient recording outcome used is psychometrically dependable and
valid in the target population.(44)

While some
missing data in longitudinal studies is nearly unavoidable, there are ways to
minimize it. The first step is literally, the first step: in the design.
Whether or not the patient record outcomes are a primary or secondary outcome,
they need to be completely integrated into the design and the manner of the
study, with this integration codified in the study protocol, quality assurance
measures, and statistical analysis plan. One design decision that impacts on
missing data is the decision to continue assessments after the patient misses
an assessment or goes off treatment. In some settings the effectiveness of
treatment on the outcome will occur after treatment failure. Continued
assessment is conservative; if it is decided later that these data are not closely
connected to the research question they can be excluded. This testament is
balanced with a warning about the length of follow-up in populations with high
rates of morbidity. Assessments should generally not be planned after the
median survival (and possibly should be a shorter interval).(44)













I'm Simon!

Would you like to get a custom essay? How about receiving a customized one?

Check it out