Research Designs

The guideline below is a recommended outline only and does not substitute for your assignment grading rubric; your paper will be graded using the assignment grading rubric criteria.  

For the introduction paragraph section, summarize your learning  using mostly your own words  (see the grading rubric for details):

· The need for nursing research.

· The importance for nurses to understand the basic principles of research.

· The purpose of your paper.

For the quantitative research section, summarize your learning  using mostly your own words  (see the grading rubric for details):

· The importance of quantitative research.

· One type of quantitative design; explain one important feature of this type of design.

· How quantitative research can help improve nursing practice.

For the qualitative research section, your learning  using mostly your own words  (see the grading rubric for details):

· The importance of qualitative research.

· One type of qualitative design; explain one important feature of this type of design.

· How qualitative research can help improve nursing practice.

For the research sampling section, your learning  using mostly your own words  (see the grading rubric for details):

· What is sampling and why is sampling important.

· One sampling strategy used in quantitative research.

· One other sampling strategy that you learned.

For the credible nursing practice section, your learning  using mostly your own words  (see the grading rubric for details):

· How research can help to make nursing practice safer.

· Why research is critical for creating an evidence-based nursing practice.

For the conclusion section, summarize your learning  using mostly your own words  (see the grading rubric for details):

· Short, concise, thorough summary of the main points of the paper.

Houser, J. (2018). Nursing research: reading, using, and creating evidence (4th ed.). Jones & Bartlett Learning.

Chapter 2: The Research Process and Ways of Knowing

CHAPTER OBJECTIVES

The study of this chapter will help the learner to

Discuss the philosophical orientations that influence the choice of a research design.

Contrast the characteristics of quantitative and qualitative research.

Review the steps involved in the research process.

Determine the way that a design is linked to the research question.

Classify research based on characteristics related to intent, type, and time.

Evaluate which kind of evidence is best provided by quantitative and qualitative research.

KEY TERMS

Applied research

Basic research

Cross-sectional methods

Experimental research

Longitudinal studies

Mixed methods

Paradigm

Prospective studies

Qualitative research

Quantitative research

Quasi-experimental studies

Retrospective studies

Chapter 6: Selecting an Appropriate Research Design

CHAPTER OBJECTIVES

The study of this chapter will help the learner to

Establish the link between the research question and the study design.

Evaluate the characteristics that are the basis for design decisions.

Differentiate the kinds of questions that require quantitative, qualitative, and mixed method designs.

Identify the types of variables that reflect the concepts in a research question.

Review designs that describe populations, test relationships, or examine causality.

Relate the type of design to the strength of evidence it can support.

KEY TERMS

Confirmatory studies

Correlation research

Dependent variable

Descriptive studies

Descriptive variables

Exploratory studies

Extraneous variables

Independent variable

Predictive research

Research design

Variable

Chapter 7: The Sampling Strategy

CHAPTER OBJECTIVES

The study of this chapter will help the learner to

Define a population and discuss the rationale for sampling.

Contrast probability sampling with nonprobability sampling.

Discuss sampling options and select an appropriate strategy.

Describe methods for estimating necessary sample size.

Discuss methods for avoiding selection bias.

Appraise how the sampling method affects research as evidence.

KEY TERMS

Convenience sampling

Ecological validity

Effect size

Exclusion criteria

External validity

Inclusion criteria

Independence

Population

Population validity

Power

Probability or random sampling

Purposeful selection

Random selection

Sample

Sampling error

Sampling frame

Selection bias

Snowball sampling (referral sampling, respondent-driven sampling)

Unit of analysis

Chapter 8: Measurement and Data~Collection

CHAPTER OBJECTIVES

The study of this chapter will help the learner to

Discuss the link between the research question and the measurement strategy.

Describe the types of reliability and validity and explain how they are assessed.

Evaluate sources of measurement error and plan strategies to minimize their effects.

Compare the advantages and disadvantages of data collection methods.

Discuss the importance of having clearly prescribed data management procedures.

Determine how the measurement strategy supports application of the data to evidence-based practice.

KEY TERMS:

Calibration

Closed questions

Codebook

Conceptual definition

Guttman scale

Internal reliability

Inter-rater reliability

Likert scale

Measurement

Measurement error

Open-ended questions

Operational definition

Photovoice

Precision

Primary data

Psychometric instruments

Random error

Responsiveness

Scales

Secondary data

Sensitivity

Specificity

Systematic error

Test blueprint

Validity

Visual analog scale (VAS)

Understanding Levels of Measurement

The initial, and perhaps most vital, step in descriptive analysis is to identify the level of measurement for each variable so as to choose the appropriate statistical analysis. This decision is the responsibility of the researchers who create descriptive studies and is an important point for critique by nurses who read research reports. Data can be collected in one of four possible levels of measurement: nominal, ordinal, interval, or ratio. Each level has characteristics that make it unique, and each requires a particular type of statistical technique. Table 11.1 shows descriptive statistical techniques that are appropriate for each level of measurement.

Nominal-level data are those that denote categories and have no rank order; numbers given to these data are strictly for showing membership in a category and are not subject to mathematical calculations. Nominal data can be counted, but are not measured, so they can be summarized using statistics that represent counts. Fall precautions is an example of nominal data: Either a patient is on fall precautions or the patient is not. Summary statistics appropriate for this level of measurement are frequency, percentage, rates, ratios, and mode.

Ordinal data are also categories but have an added characteristic of rank order. These data differ from nominal data in that the categories for a variable can be identified as being less than or greater than one another. However, because the level of measure is still categorical, the exact level of difference cannot be identified. A pain scale, in spite of its representation as a series of numbers, is an example of ordinal data. For example, while we know a score of 7 on this scale is greater than a 5, the difference cannot be quantified. For example, we cannot conclude that the difference between a 7 and a 5 on the pain scale is the same as the difference between a 4 and a 2 on the same scale; we simply know that one number is higher than the other. Further, we do not know that a score of 5 for one patient is the same as a score of 5 for another patient. The fact that the pain scale is represented by numbers also does not necessarily mean the characteristic under study has been quantified. In our example, we do not know exactly how much pain exists; it is just rated against other experiences of pain for that patient. The patient’s score has simply been ranked against all other values of the variable. Statistical techniques appropriate for ordinal data include those appropriate for nominal data plus range, median, minimum, and maximum.

Interval and ratio data are recorded on a continuous scale that has equal intervals between all entries; length of stay is an example. Data collected on interval or ratio levels result in numbers that can be subjected to many mathematical procedures, including mean, standard deviation, variance, and evaluation of the distribution (skew and kurtosis).

Identifying Shape and Distribution

Initial analyses of data are meant to help the researcher identify the distribution, and therefore the shape, of the variable’s data. The outcome of this analysis, coupled with the level of measurement, guides the researcher in selecting the appropriate statistics to represent the variable’s center and spread.

Summarizing Data Using Frequencies

Frequency is a statistical term that means a count of the instances in which a number or category occurs in a data set. Frequencies are commonly used in clinical settings; for example, a frequency might be used to document the number of infections by surgery type, the number of patient falls by nursing unit, or the number of nurses who leave in the first 18 months of employment. In research, frequencies are used to count the number of times that a variable has a particular value or score. A researcher may collect data on nominal-level variables or ordinal-level variables and then generate a frequency count per category to summarize the data. For example, if information about gender were desired, two values (male and female) would be collected; the number of participants in the study who were male and the number who were female would be tallied.

Frequency data can also be used to calculate percentages, rates, and derived variables. A percentage is a useful summary technique that shows the relative frequency of a variable. For example, if gender was measured as a variable and there were 180 female participants in a sample of 400, the percentage of female participants would be 45%. The number 45% is more meaningful as a summary value than the frequency count because readers can tell quickly that slightly less than half of the sample was female. To calculate the percentage in this example, the number of female participants is divided by the number of participants in the entire sample (180/400 = 0.45 or 45%).

Rates that are clinically important can be calculated to provide information about data trends over time. Like a percentage, a rate is calculated by dividing the frequency of an event in a given time period by all possible occurrences of the event during the same time period. The difference is that percentages by definition are “per 100,” whereas rates can have a different denominator, such as per 1000 patient-days. Monthly fall rates are an example; the number of falls in a month is divided by the total number of patient-days in that month and then multiplied by 1000 to give the number of falls per 1000 patient-days. This allows for comparison between units based on opportunities for falls (each day) instead of the raw count (number of falls). When calculated periodically, patient outcomes expressed as rates can be monitored as a basis for action planning to improve care (Altman, 2006).