# Reading Research Articles how to approach statistical information

Document Sample

```					 Reading Research Articles
how to approach statistical information

Marion Haas
March 18, 2008
Introduction
   In this session…
   Aim
   Practical skills to learn
   Type of statistical information presented
   Practice!
Aim of this session
   Recognise diversity among students
   Often not mathematical
   Different notions of statistics
   Translate concepts into understandable language
   Approach levels of statistics in a step-wise way
   Knowing where to look in the text
   Interpreting the text
   Explaining the text in your words
Definition
Statistics is a mathematical science pertaining to
the collection, analysis, interpretation or
explanation, and presentation of data. It is
applicable to a wide variety of academic
disciplines, from the natural and social
sciences to the humanities. Statistics is also
used for making informed decisions in
Statistics
   Often first contact with numbers beyond
simple concepts at school
   Different way of using numbers
   All future researchers & professionals in many
areas need:
   Working knowledge of how to understand
statistical output
   Some sophisticated understanding
Approach (1)
   Develop and broad & integrated view of
statistics
   Use the text – know where to look
   Interpret the text – what does it mean?
   Explain the text – use your words
   Ask a series of questions
   Understand the statistical content of articles
   Connecting topic to professional lives
Approach (2)
   3 types articles
   Elementary descriptive statistics techniques
   Introductory techniques of statistical inference
   Preliminary: the same for all articles
   In-depth: higher order skills used to
   Generate text & analyses
   Place text in context
1.   General skim to decide if useful
   Title, author, abstract, first & last paragraphs,
diagrams, graphs, tables
   5 minutes, 5 dot points
2.   If continuing
   Questions about aim, audience, context
   Understand main points of article
Specific preliminary questions (1)
1.   What is the research question?
   How is it presented?
   Why is it presented in this way?

2.   What research methods are used?
   Observational, experimental, neither
   Sampling techniques?
Specific preliminary questions (2)
3.   What data are used?
   How are they dealt with?
   How are they presented/described?

4.   Statistical techniques?
   List
   Why is each one used?
Applied statistics
   Descriptive statistics
   Statistical methods used to summarize or describe
a collection of data
   Inferential statistics.
   Modeling patterns in the data to account for
randomness and uncertainty in the observations, &
used to draw inferences about the process or
population being studied
   Both descriptive and inferential statistics
comprise applied statistics
Practice
   Let’s look at the first article
   Veal, Tony (1997). Gambling trends in the 1990s.

   Skim for 5 minutes
   First and last paragraphs
   Diagrams, graphs, tables

   What sort of statistical “work” (activity) has Veal
undertaken?
   What type of language has he used?
Descriptive statistics
   Used to summarise data
   Numerically or graphically
   Describe the sample
   Numerical
   Mean, median, mode, range, standard deviation etc
   Graphical
   Charts and graphs
Veal in depth
1.       Aim and audience
     What is the main aim
     What clues do the statistical techniques give about the
aim?
     What audience is it written for?
2.       Content
     Figure 1: In 1996, which was the largest component of
gambling expenditure?
     Which was growing fastest?
     Why is NSW the “most mature” gambling market?
Veal (2)
3.   Analysis
   Do the graphics work here? Why?
   How could they be improved?
   Is the information presented in an objective
fashion? Does the author approve of gambling?
   Is the writing style formal or informal?
Examples?
Inferential statistics
   Used to:
   Model patterns in data
   Account for randomness
   Draw inferences about the larger population
   Inferences
   Make estimates of numerical characteristics (estimation)
   Describe associations (correlation)
   Model relationships (regression, ANOVA, time series)
Important concepts (1)
   Correlation
   When two characteristics tend to vary together, as
if connected
   Eg income and age at death
   Poorer people die younger
   BUT, does this mean that poverty causes death or that
poor health causes poverty?
   Correlation DOES NOT imply causation
Important concepts (2)
   Probability
   Fundamental concept used to understand randomness
   Methods able to estimate and correct for randomness in
design, sample, data collection
   Alpha level refers to probability of Type 1 error
   Significant association found when does not exist
   Alpha level conventionally set at 0.05 ie if p=<0.05 we accept that
a significant association exists
significantly different if DO NOT overlap
Statistical methods
   What will happen to response (outcome,
dependent) variables if changes occur in predictor
(independent) variables?
   Experimental & observational studies
   Both investigate how changing independent
variables affects behaviour of dependent
variable(s)
   Difference lies in HOW the study is conducted
Experimental study
   Basic design
   Initial measurement
   Change conditions
   Measure again

   Control group
   Blind measurement
Observational study
   Basic design
   Data gathered
   Correlations between predictors (independent) and
response (dependent) variables investigated

   Control
Practice
   General questions
   Viney et al: Skim for 5 minutes
   What is the research question
   What research methods are used?
   What data are used?
   What statistical techniques are used?
In-depth questions (1)
   Aim and audience
   What is the main aim?
   What clues do the statistical techniques give about
the aim?
   What audience is this aimed at? Reasons?
   Content
   What is a randomised trial?
   What does “eligibility criteria” mean?
In-depth questions (2)
   Analysis
   Who were the subjects for this trial? how were
they selected?
   “there was no significant difference between the
arms in the number of thoracotomies avoided
(Table 3, p=0.2)”. Can you explain this in lay
terms?
   Can you think of some reasons that the
management of NSCLC did not change as much as
it could have?
More practice
   Vignaendra and Fitzgerald
   What is the research question
   What research methods are used?
   What data are used?
   What statistical techniques are used?
In-depth questions (1)
   Aim and audience
   What is the main aim?
   What clues do the statistical techniques give about
the aim?
   What audience is this aimed at? Reasons?
In-depth questions (2)
   Content
   What do the authors mean by “caution cohort” and
“conference cohort”?
   Explain what is meant by “bi-variate” and “multi-
variate” analysis
   Table 2: what characteristics (variables) are NOT
significant? Why is this important?
In-depth questions (3)
   Analysis
   Look at the logistic regression model/s (Tables 2
and 6)
   What is the response (dependent) variable?
   What are the predictor (independent) variables?
   Can you explain these results in lay terms
   What does survival analysis mean in this context?
   What % of young people “survived” at least 12 months
before being seen in court again? (Figures 3 & 4)
   By what time following the first conference had 50% of
young people been seen in court again? (Figure 4)
Reference
Leigh Wood and Peter Petocz (2003). Reading
Statistics, Mathematics Study Centre,
University of Technology, Sydney. Printed by
UTS Printing Services

```
DOCUMENT INFO
Shared By:
Categories:
Stats:
 views: 15 posted: 11/14/2008 language: English pages: 29
How are you planning on using Docstoc?