Introduction to Statistics

• Statistics describes either statistical data or statistical methods.
• Statistics involves the study and application of theories and methods for data analysis, especially data from random events.
• The field provides tools and techniques for the scientific method, including:
• Forming hypotheses
• Designing experiments and observational studies
• Gathering data
• Summarizing data
• Drawing inferences from data e.g. testing hypotheses.
• In addition to the field of statistics, the term 'statistic' can also refer to a numerical quantity derived from sample data, such as mean, median, or maximum.
• Statistics has 2 branches Mathematical and Applied Statistics.
• Mathematical Statistics is concerned with the study and development of statistical theory/methods.
• Applied Statistics applies statistical methods to solve real-world problems using randomly generated data and develops new methodologies when needed.
• Biostatistics applies statistics to health sciences and biology.
• Biometry differs from biostatistics depending on the application focus: health sciences for biostatistics, broader biology (agriculture, ecology, wildlife biology) for biometry.

Why Study Biostatistics?

• Statistical methods are commonly used in health applications, like survival analysis and longitudinal data analysis.
• It makes the subject more engaging to those in health-related careers.
• Methodology is applicable to similar problems encountered in real life.

Characteristics of Statistical Data

• For numerical descriptions to be considered statistics, they must be aggregates of facts, not single, isolated facts.
• Statistics should be influenced by multiple factors. For instance, an outbreak is due to human, parasite, mosquito, and environmental factors.
• Data must be enumerated or estimated with reasonable accuracy to be reliable for statistical investigations.
• Collection should be systematic and for a specific purpose that is clear to the enumerator to avoid confusion.
• Data must be comparable and placed in relation to each other, whether by time, space, or condition, to be useful.

Rationale for Studying Statistics

• Statistics provides a formal framework for organizing information, rather than relying on anecdotes or personal experiences.
• Quantitative measurements are increasing in medicine and public health.
• Biological processes inherently have significant variation.
• Public health and medicine are becoming more quantitative as physicians encounter more quantitative information due to technological advancements.
• Statistics helps assemble and handle quantitative data, vital for interpreting results.
• Medical research increasingly relies on statistical technology for planning, conducting, and interpreting.
• Statistical analysis addresses critical questions, for example:
• The effectiveness of new treatments.
• Potential side effects
• Appropriate sample sizes
• Normal clinical measurement ranges.
• Reliability and validity of measurements, and the impact of laboratory errors.
• Statistics is pervasive in medical literature due to the increasing quantitative nature of public health and medicine.
• Interpreting data in the presence of variability is central to statistics.

Relevance of Statistical Literacy in Nursing

• Nursing uses scientific research to achieve positive patient outcomes, making statistics vital.
• It directly impacts patient care and can change policies on a larger scale.
• Numbers in healthcare influence and inform, requiring nurses to understand and critically analyze them.
• Statistical knowledge and skills are essential for nurses.

How Statistics Improve Nursing

• Overcoming misconceptions requires statistical knowledge and skills.
• Statistics has been fundamental in nursing since Florence Nightingale, the first nurse statistician who used it to show that better sanitation reduced deaths during the Crimean War.
• Nightingale focused on standardizing medical data collection and benchmarking.
• Nurses require statistics to advance and improve the profession.
• Clinical research increases yearly, expecting nurses to implement evidence-based practices.
• Statistics enables nurses to prioritize treatments and identify when patients need follow-up or urgent care.
• Nurses use statistics to spot trends in vital signs and symptoms to make well-informed decisions about patient care.
• Data sheets and frequency charts help in documenting medication timing, illustrating statistical use.
• It helps nurses understand patients; data can reveal issues like ICU patients experiencing hallucinations, which can quantify the importance and frequency of these issues.

Statistics Foster Evidence-Based Practice

• Nursing relies more on empirical evidence for effective patient care protocols.
• Clinicians need statistical understanding to interpret relevant literature.
• Using statistics, clinicians can revise methods based on research.
• Using statistics advances nursing, informing best practices and continuous critical care knowledge updates.
• Combining caring with statistics is not new; Florence Nightingale used statistics to improve practice and advocate for policy changes.
• Numbers can show that better sanitary conditions reduce deaths.
• Statistics helps allocate resources and bring factual change versus emotional arguments.

Educating Nursing Trainees

• Understanding statistics is crucial for trainee nurses in any program.
• Nursing students should improve skills and confidence to provide top-level care.
• They need to develop a working knowledge of statistics but do not need to be statisticians.
• Statistical literacy is needed to undertake research projects, interact with researchers/statisticians, and to design and implement projects effectively.
• Grasping imagination and quantifying perception are aided by statistics.
• Using statistics is more effective than relying on emotion or opinion to support arguments for change.
• Statistics enable nurses to assess the effectiveness of interventions, such as dressing types and glove usage, informing best practices.

Statistics Teach Caution

• Statistics highlight the complexity of human life and experiences but rarely "prove" anything definitively.
• Discovering causation may take years and repeated studies.
• It is important to see the bigger picture rather than jumping to conclusions from single findings.
• Modern statistics present a mixed blessing as they can quickly produce numerous statistics, increasing the chance of finding significant results from incorrect tests or high data volume.
• A technically perfect statistical analysis cannot fix data collection problems.
• When assessing study results, probe deeply, challenge assumptions, and avoid hasty conclusions.
• Critical care nurses can calculate drug dosages, understand body parameters, and manage complex care, investing a small effort to understand statistics.
• Statistical literacy involves distinguishing statements of association from causation, as associations are often mistaken for causal relationships.

Levels of Measurement

• "Levels of measurement" and "scales of measure" are expressions referring to Stanley Smith Stevens' theory of scale types.
• There are four levels of measurement for a variable:
• Nominal
• Ordinal
• Interval
• Ratio
• Nominal Level: Numbers classify data; words, letters, and alphanumeric symbols can also be used for categorization.
• Ordinal Level: Measurement shows a ranked relationship, indicating the order of measurements.
• Interval Level: Measurement classifies and orders data. It also specifies equal intervals along the scale, like temperature in centigrade.
• Ratio Level: In addition to equal intervals, observations can have a true zero value, indicating none of the variable, such as temperature measured in Kelvin

Sources of Data

• Data in Statistics come from statistical and non-statistical sources.
• Statistical sources include censuses and official surveys.
• Non-statistical sources come from administrative or private sector data collection.
• Statistical Survey: Conducted using a sample, known as a Sample Survey, it involves collecting and analyzing sample data to estimate population characteristics.
• Census: Unlike a sample survey, a census collects and analyzes data from all population items within a specific reference period.
• Registries: These are statistical information storehouses, which are reliable and detailed for data collection and analysis

2 Types of Data

• Data are classified as primary and secondary.
• Primary data are collected first-hand by the surveyor for a specific purpose, that has not undergone statistical treatment.
• Secondary data have already been collected and published by an and have undergone at least one round of statistical treatment.

Methods of Primary Data Collection

• Personal investigation: The surveyor collects data themselves
• Use investigators
• Questionnaires suit the study and can be mailed
• Telephonic investigation: Quick and accurate information is collected through phone

Methods of Secondary Data Collection

• Official government publications, such as those from the Ministry of Finance and Statistical Departments.
• Commerce and trade associations release data.
• Newspapers, journals and technical publications contain articles.

Statistical Variables

• Experiments examine variables, meaning what can be measured, manipulated, or controlled.
• Variables fall into 2 basic categories:
• Independent
• Dependent

Independent Variables

• Sometimes called experimental or predictor variables, are those manipulated to ascertain their impact on a dependent/outcome variable.

Experimental vs Non-Experimental Research

• Experimental Research: Aims to understand the effect of manipulating indepent variables on dependent variables.
• The potential to manipulate independent variables makes determination of cause and effect determination possible.
• Non-experimental Research: Manipulating independent variables is not possible or ethical, so researchers may not have an opportunity to do so.
• Illegal drug use (independent variable) on certain behavior can be studied by surveying drug users and non-drug users and using questionnaires.
• The association or relationship can be examined but not necessarily cause and effect.

Categories of Experimental Variables

• Categorical variables, also known as discrete or qualitative variables, can be nominal, ordinal, or dichotomous.
• Nominal variables consist of 2+ categories without an intrinsic order.
• Dichotomous variables are a specific class of nominal variables with only 2 levels, for example:
• Sex
• Gender
• Ordinal variables, like nominal variables, have 2+ categories but can be ordered or ranked.
• Continuous variables, also known as quantitative variables, are classified as either interval or ratio variables.
• Interval variables involve measurement along a continuum with numerical values where equal differences on the scale mean equal differences in what is being measured, as with temperature measured in Celsius or Fahrenheit.
• Ratio variables are interval variables with the added condition that zero indicates none of the variable.
• For example, temperature measured in Kelvin is a ratio variable because 0 Kelvin (absolute zero) means there is no temperature.
• Examples include
• Height
• Mass
• Distance