Professional Documents
Culture Documents
CAUSALITY
- Questions about diagnosis or assessment, as
well as questions about status quo of health-
related situations, are typically descriptive.
- Directly determine or examining the cause and Several writers have proposed criteria for
effect. establishing a cause-and-effect relationship.
- Defines the general concept of cause According to Lazarsfield (1995), reflecting idea of
- Most phenomenon have multiple causes John Stuart Mill, identified three criteria for
causality:
EXAMPLE:
• Do telephone therapy interventions for patients 1.TEMPORAL
diagnose with prostate cancer cause - A cause must precede an effect in time. If we
improvements in their decision-making skills? were testing the hypothesis that aspartame
(Intervention prognosis) causes fetal abnormalities, it would be
• Do birth weight under 1,500 grams cause necessary to demonstrate that the
development delays in children? (Prognosis abnormalities did not develop before the
question) mothers’ exposure to aspartame.
• Does cigarette smoking cause lung cancer
(Etiology/Harm question) 2. EMPIRICAL RELATIONSHIP
- There be an empirical relationship between
EXAMPLE OF GENERAL CONCEPT OF THE CAUSE the presumed cause and the presumed
• We understand failure to sleep causes fatigue effect.
and that high-caloric make causes weight gain. - In the aspartame example, we would have
• Weight gain can be the effect of high-caloric to find an association between aspartame
consumption. consumption and fetal abnormalities, that is,
that higher percentage of aspartame users
1|P a ge
than nonusers had infants with fetal - Examination of the effect of independent
abnormalities. variable on dependent variable
- A basic distinction in your quantitative research
3. CAUSED BY A THIRD VARIABLE design is between your experimental and non-
- The relationship cannot be explained as experimental.
being caused by a variable. Suppose, for - In an experimental (or randomized controlled
instance, that people who used aspartame trial, RCT), randomization is there. Controlled
tended also to drink more coffee than criteria is there. In here, there is an experimental
nonusers of aspartame. group and controlled group. When we say
experimental group, yan ang binibigyan ng
There would then be a possibility that any intervention or treatment. When we say the
relationship between maternal aspartame use and controlled group, walang ibibgay na treatment
fetal abnormalities reflects an underlying causal diyan.
relationship between a substance in coffee and the - Random – Random picked
abnormalities.
TRUE EXPERIMENTAL DESIGN
Additional criteria were proposed by - Researcher have complete control over the
Bradford-Hill (1971) as part of the discussion about extraneous variable and can predict confidently
the causal link between smoking and lung cancer. that the observed effect on the dependent
Two of Bradford-Hill’s criteria foreshadow the variable is only due to the manipulation of t
importance of meta-analyses, techniques for which variable.
had not been fully developed when the criteria were
proposed. The criterion of coherence involves similar CHARACTERISTICS OF A TRUE EXPERIMENTAL
evidence from multiple sources, and the criterion of DESIGN
consistency involves having similar levels of statistical Manipulation
relationship in several studies. Another important - it pertains to your experimental or intervening
criterion in health research is biologic plausibility, group.
that is, evidence from basic physiologic studies that - Conscious control of independent variable.
a causal pathway is credible.
Control
Researchers investigating casual - that is the control group (hindi binibigyan ng
relationships must provide persuasive evidence interventions)
regarding these criteria through their research
design. Some designs are better at revealing cause- Randomization
and-effect relationships than others, but not all - everybody has an equal chance to be picked
research questions can be answered using the - every subject has an equal chance of
strongest designs because of ethical or practical assigned to experimental or control group.
constraints.
Method
CLASSIFICATION Flip a coin
I. Experimental Research Design Random table
- True experimental design Drawing lots
- Quasi experimental design Fish bowl method
- Pre-experimental design
2|P a ge
4. Factorial design Based on the example:
5. Cross over design ● Subject – standard five children
● Treatment- smoking prevention programme.
FACTORIAL DESIGN
- Researcher manipulates 2 or more independent
variables simultaneously to observe their effect
on the dependent variable.
- Useful when more than two independents
variables to be tested
- Testing of several hypothesis at a single time
EXAMPLE:
Effect of two different protocols of mouth care on
prevention of VAP when performed at different
EXAMPLE: frequencies in a day.
A study to assess the effect of an educational
intervention related to urinary incontinence on the
subsequent help seeking of older adults.
EXAMPLE:
Effect of three different antihypertensive drugs on
patient with hypertension
3|P a ge
- Identical to pretest post-test control group
EXAMPLE: design, except no random assignment of
Effectiveness of chlorhexidine and saline mouth care subjects
protocol on prevention of VAP
EXAMPLE:
To study the effects of integrated care on equality of
work in nursing homes.
ADVANTAGES:
• Most powerful design to establish the causal
relationship between the independent and
dependent variable.
• Controlled environment in which study is
conducted can yield a greater degree of purity
in observation.
• Controlled not found in natural setting can be
created in an experimental setting Based on the example:
• Researcher can pursue studies in more leisurely, • Independent variable –quality of work in nursing
careful and concentrated way. homes
• Treatment- integrated care
DISADVANTAGES:
• Results cannot be replicated in studies TIME SERIES DESIGN
conducted on human due to ethical problems.
- When experimenter wants to measure the
• It is not possible to impose control over effects of a treatment over a long period of time.
extraneous variable Experimenter continue to administer the
• Difficult to get cooperation from the study treatment and measure the effects a number
participants
• Many of human variables neither have valid
measurable criteria nor instrument to measure.
EXAMPLE:
Quasi-independent variable are used instead
Measuring a child's school performance on a weekly
of true independent variable
basis and then introducing a new teaching
technique. Then again measuring on a weekly basis.
TYPE
1. Non-randomized control group design ADVANTAGES:
2. Time series design • Practical and feasible to conduct
• More suitable for real world natural setting than
NON-RANDOMIZED CONTROL GROUP DESIGN true experimental design
- Non-equivalent control group design
4|P a ge
• Allow researcher to evaluate the impact of experimental or quasi-experimental design.
quasi-independent variable under naturally
occurring conditions For example, earlier we posed this prognosis
• May be able to establish causal relationship. question: Do birth weights under 1,500 grams cause
developmental delays in children? Clearly, we
DISADVANTAGES: cannot manipulate birth weight, the independent
• No control over extraneous variables variable. Babies’ weights are neither random nor
• Absence of a control group make the results less subject to research control. One Way to answer this
reliable and weak for establishment of the causal question is to compare two groups of infants – babies
relationship between independent and with birth weights above and below 1500 grams at
dependent variables. birth—in terms of their subsequent development.
7|P a ge
SURVEY RESEARCH DESIGN 3. Secondary data analysis
- Used to collect information from different subject 4. Outcome research
within a given population having same 5. Evaluation studies
characteristics of interest. 6. Operational research
OPERATIONAL RESEARCH
- Involves study of complex human organizations
and services to develop new knowledge about
institutions, programmes, use of facilities and
personnel in order to improve working efficiency
of an organization.
9|P a ge