Sampling
| The subset of the population used to study the population as a whole. |
| What is a sample? |
The way in which a variable is operationally defined, resulting in data that are categorical, rank-ordered, or continuous (numeric). |
What is the level of measurement? |
| Confidence that an observed effect in a dependent variable was actually caused by the independent variable and not other factors. |
| What is internal validity? |
| Statistics that describe or summarize the characteristics of a given variable, including its central tendency and dispersion. |
| What are descriptive statistics? |
| An approach to practice that integrates knowledge of the best available empirical evidence for practice with relevant client attributes and the practitioner's expertise. |
| What is Evidence-Based Practice? |
| Sampling that begins with a predetermined sampling frame and depends on chance. Forms include simple random sampling, stratified sampling, systematic sampling, and cluster sampling. |
| What is probability-based sampling? |
| A measurement tool that yields consistent scores. This might be demonstrated through adequate Internal Consistency, Test-Retest, or Interrater approaches. |
| What is a reliable measure? |
| Factors that impede the ability to conclude that an effect observed in the dependent variable was caused by the independent variable. These include History, Instrumentation, Maturation, Testing, Statistical Regression, and Selection Biases. |
| What are the design-related threats to internal validity? |
| Descriptive statistics reporting central tendency and dispersion for nominal- and sometimes ordinal-level variables. |
| What are frequency and percentage? |
| A reference tool, reflecting the relative authority of various types of empirical studies, used in determining the "best available evidence for practice". |
| What is the evidence hierarchy? |
| Sampling that does not rely on chance or randomness. Forms include availability/convenience sampling, purposive sampling, quota sampling, and snowball sampling. |
| What is non probability-based sampling? |
| A measurement tool that measures what it is believed to measure. This might be demonstrated through subjective assessment by experts or by comparison to another tool that is known to measure the same concept. |
| What is a valid measure? |
| A strong research design that includes random assignment of study participants to different intervention groups. |
| What is an experimental design? |
| Common descriptive statistic used to report central tendency for interval/ratio-level variables? (Bonus: which sister statistic reports dispersion for variables at this level of measurement?) |
| What is the mean? (And...cha ching, bonus points: What is the standard deviation?) |
| Exploration, Description, Explanation, and Evaluation. |
| What are the major purposes of research? |
| The theory that is relied on for producing study findings that are safely generalizable to the population from which the sample was drawn. |
| What is probability theory? |
| Problems in measurement - either systematic (as with bias) or random - that result in inaccurate data. |
| What is measurement error? |
| A research design that includes some control for threats to internal validity through inclusion of a (nonequivalent) comparison group or extended collection of pretest data. |
| What is a quasi-experimental design? |
| Statistical tests that are used for making inferences based on sample-based findings to some larger population. |
| What are inferential statistics? |
| A rating provided by Journal Citation Reports, depicting the relative importance of a journal within its field. |
| What is a Journal Impact Factor? |
| The extent to which the relationships among variables found in a study can be generalized beyond that particular study? |
| What is external validity? |
| Known scales, subject-supplied data, observation, archival data, and biological/physiological data. |
| What are measurement techniques or approaches? |
| A weak research design that has minimal or no control for threats to internal validity. |
| What is a pre-experimental or non-experimental design? |
| A very important reference statistic; represents the likelihood that an observed empirical relationship is simply a chance occurrence. |
| What is the p-value (aka level of significance)? |
| A series of helpful Evidence-Based Practice reference pages hosted by the USC Libraries, including specific social work pages. |
| What are the social work LibGuides? |