This document outlines aspects of interpreting quantitative research results. It discusses interpreting results with graphs and diagrams, credibility and different types of biases, magnitude and precision of results, and clinical versus statistical significance. It provides examples of interpreting hypothesized, non-significant, and unhypothesized results. The document emphasizes considering validity, bias, corroboration, and effect sizes when interpreting results as well as implications, generalizability, and significance of findings.
5. Interpretation Vs Inference
Credibility & Validity
Credibility & Bias
Credibility & Corroboration
Magnitude & Precision of Results
Aspects of Interpretation
Multiple Inferences
Conclusion
from Singles
inference Inference
Interpretation
Do the findings represent the whole population?
“Truth in the real World”
6. Credibility & Validity
Desired inferences are
valid in terms of
Sampling
Intervention Design
Measurement
Research design
Analysis
Effectiveness of on Intervention Package(IP) to increase the
physical activity among low income group women in
California.
Construct Validity – Population Construct
Statistical conclusion Validity- Power Analysis
External Validity- Sampling
Internal Validity – Sample composition, Attrition
7. Credibility & Bias
Research
Design
Sampling Measurement Analysis
Expectation
Bias
Sampling
Error
Social
desirability
Bias
Type- I Error
Hawthorne
Effect
Volunteer
Bias
Extreme
Response Bias
Type- II Error
Contamination
of Treatment
Non response
Bias
Reactivity
Non
compliance
Bias
Observer Bias
Selection Bias
Attrition Bias
History Bias
Credibility & Corroboration
Triangulation
Consistency
Replication
9. Meaning of the Quantitative Results
Research Question
Does ECT causes ECT? Does the
intervention has impact on head ache?
Research results
59.4%(95% CI [56.3,63.1]
Interpretation
59.4% of patients in both group
experienced head ache(%) and the ECT
causes the head ache(CI)
Compare the effectiveness of Acetaminophen Vs Cryotherapy on
Headache induced by ECT among patients undergoing ECT
Research Question
Does the prevalence of head ache
differs between the intervention
groups
[Cryotherapy/Acetaminophen]]?
Interpretation??
Hypothesis Testing
10. Statistical Significance
Compare the probability from test (the p-value) to the critical probability value that
was hypothesized (the alpha level)
P-Value: the probability that the results
were due to chance and not based on
intervention. P-values range from 0 to
1. The lower the p-value, the more
likely it is that a difference occurred as a
result of intervention.
Alpha (α) level: Alpha is often set at .05 or .01.
The alpha level is also known as the Type I
error . An alpha of .05 means that can accept
that there is a 5% chance that the results are
due to chance rather than intervention.
If the p-value is less than the alpha value, conclude that the difference observed is statistically significant
11. Hypothesis Testing & Interpretation
A study was to examine the relationship between behavior problems and self-concept in
children and adolescents with ADHD. The relationships among gender, age, and behavior
problems with self-concept were explored, and ethnic differences with respect to behavior
problems and self-concept were examined. Houck, G., Kendall, J., Miller, A., Morrell, P., & Wiebe, G.
(2011). Self-concept in children and adolescents with attention deficit hyperactivity disorder. Journal of pediatric
nursing, 26(3), 239–247. https://doi.org/10.1016/j.pedn.2010.02.004
12. • Behavior problems
in ADHD children is
associated with low
self concept
Hypothesis
• Internalizing
problems were
significantly
predictive of lower
self concept scores
Interpretation
• Age and
internalizing
behavior is found to
be negatively
influence the child’s
self concept
Discussion
Hypothesis Testing & Interpretation
13. Interpreting - Non hypothesized Results
Non significant results pose interpretive challenges
Statistical test are geared to disconfirmation of Null Hypothesis
Failure to reject null hypothesis(Type II Error) may result from
• Small sample size
• Unreliable measures
• Poor internal validity
• Anomalous sample
But when the actual research hypothesis is null (prediction of NO group
difference) stringent additional strategies must be used to provide
supporting evidence .
It is useful to compute effect size or is to illustrate that the risk of Type II
error is small.
14. 1. Serendipitous significant Findings
Exploring relationships that were not
considered during the design of the study
Eg. Latendresse and Ruiz (2011) studied the
relationship between chronic maternal stress
and preterm birth.
They reported an unexpected finding that
maternal use of selective serotonin reuptake
inhibitors(SSRIs) was associated with 12 fold
increase in preterm births
Interpreting Un Hypothesised Results
2. Significant Result contrary to Hypothesis
Obtaining results opposite to those
hypothesized
Eg. Dotson & Collegues (2014) , who tested
hypotheses about nurse retention with a
Sample of 861 registered nurses (RNs) predicted
that higher levels of altruism would be
associated with stronger intentions to stay in
Nursing , however , the opposite was found.
They speculated that this might mean that
some nurses “ are no long experiencing the
fulfillment of their altruistic desires in the field
of Nursing
Un hypothesized significant results can occur in
15. Generalizability of Results Implication
Which Population, setting,
versions of intervention were
the study operations good
Consider proxy
Are these results relevant to
any particular situation?
A Study that have limited
credibility or importance
may have little utility in
clinical practice.
16. The practical importance of
research results in terms of
Genuinely
Palpable effect on daily lives of patients and
health care decisions
C l i n i c a l S i g n i f i c a n c e
No change over time
Implication for health
management
Patient’s view(quality of Life)
Point in time outcomes
Change scores
Standardization possibilities
17. Clinical Significance – Group Level
Involves using statistical significance other that P value
The most widely group level indices for clinical significance at group
level used are
Effect size indexes- Magnitude of the result
CIs- Precision of the results
Number need to Treat(NNT)- Patient outcome
C l i n i c a l S i g n i f i c a n c e
18. Clinical Significance – Individual Level
Benchmark/ Threshold
Designates the score value on a measure (for the value of
the change score0 that would be considered clinically
important
With an established benchmark for clinical significance ,
each person in a study can be classified
As having or not having a score or change score that is
clinically significant.
C l i n i c a l S i g n i f i c a n c e
19. Calculate Effect size for given values and
interpret
Example: Statistical significance vs clinical significance large study compared two weight loss
methods with 13,000 participants in a control intervention group and 13,000 participants in an
experimental intervention group. The control intervention used scientifically backed methods for
weight loss, while the experimental intervention group used a new app-based method.
After six months, the mean weight loss (kg) for the experimental intervention group (M =
10.6, SD = 6.7) was marginally higher than the mean weight loss for the control intervention
group (M = 10.5, SD = 6.8).
Cohen’s d formula Explanation
•x̄1= mean of Group 1
•x̄2= mean of Group 2
•s = standard deviation
d = (x̄1 − x̄2) ÷ s
d = (10.6 − 10.5 )÷ 6.8 = 0.015
With a Cohen’s d of 0.015, there’s limited to no practical significance of the finding that the experimental
intervention was more successful than the control intervention.
22. Journal References
1. Dilip Kumar Kulkarni 2016 Indian Journal of Anesthesia , Interpretation
and display of research results Indian J Anaesth. Sep; 60(9): 657–661.
doi: 10.4103/0019-5049.190622 PMCID: PMC5037947 PMID: 27729693
2. Michael J. Albers Quantitative Data Analysis—In the Graduate
Curriculum February 26, 2017 Journal of Technical Writing and
Communication
3. The Data Puzzle: The Nature of Interpretation in Quantitative Research ,
American Journal of Political science, Vol. 40, No. 1 (Feb., 1996), pp. 1-32 (32
pages)
Published By: Midwest Political Science Association
23. Book References
POLIT & BECK Canadian Nurses Understanding Nursing Research 6th edition
(2012) Lippincott Publication
Judith Haber- Methods of critical appraisal – Nursing Research 12th edition(2018)
Wolter Kluwer Publications
Janey Hawker Nursing Research Critical Apraisal and Underdtanding 4th
edition(2004) Wolter Kleuwer Publications
Polit & Beck South Asian Edition of Nursing Research- A critical First Edition
Understanding (2019) Wolter Kluer Publications