TLDR;
This video explains the fundamental concepts of variables, measurement scales, sampling methods, and experimental design in statistics. It covers the differences between qualitative and quantitative variables, discrete and continuous variables, and the four levels of measurement: nominal, ordinal, interval, and ratio. The video also discusses various sampling techniques, including simple random sampling, stratified sampling, systematic sampling, cluster sampling, and convenience sampling, as well as potential biases in study design, such as confounding variables and overgeneralization.
- Qualitative vs. Quantitative Variables
- Discrete vs. Continuous Variables
- Nominal, Ordinal, Interval, and Ratio Scales
- Sampling Methods and Experimental Design
Qualitative and Quantitative Variables [0:04]
The video introduces two main types of variables: qualitative (or categorical) and quantitative. Qualitative variables categorize data by names or labels (e.g., hair color), while quantitative variables represent numerical data. Quantitative variables are further divided into discrete and continuous variables. Discrete variables can only take specific, countable values (e.g., number of fleas on a cat), whereas continuous variables can take any value within a range (e.g., height, weight). Time is considered a continuous variable due to the infinite possibilities between any two points in time.
Discrete vs. Continuous Variables: Examples [2:26]
The video provides examples to differentiate between discrete and continuous variables. IQ scores and ACT scores are discrete because they are specific, countable values. Height and commuting distance are continuous because they can take on infinite values within a range. Shoe size is discrete because it only comes in specific sizes (e.g., 8, 8.5, 9) without infinite gradations in between. The weight of a cat is continuous (measured), while the number of fleas on a cat is discrete (counted).
Measurement Scales: Nominal, Ordinal, Interval, and Ratio [6:30]
The video explains four measurement scales used to categorize variables based on their characteristics. Nominal scales categorize data by name or category without any order (e.g., gender, hair color). Ordinal scales categorize data with a specific order or ranking (e.g., course grades, sizes of drinks). Interval scales are quantitative and have an order with meaningful differences between values, but zero does not indicate the absence of the quantity (e.g., temperature in Fahrenheit). Ratio scales have all the properties of interval scales, but zero indicates the absence of the quantity, and ratios between values are meaningful (e.g., height, weight, distance). Nominal and ordinal scales are used for qualitative variables, while interval and ratio scales are used for quantitative variables.
Measurement Scales: Examples [11:48]
The video provides examples to illustrate the four measurement scales. Gender is nominal because it is a category without order. IQ is interval because the difference between scores is meaningful, but zero does not mean the absence of intelligence. Commuting distance is ratio because zero means no distance, and ratios are meaningful. Pain rating is ordinal because the values have a specific order, but the differences between them are not consistent. Time on a clock is interval because zero has a meaning (midnight), and length of time to take a rest is ratio because zero means no time was taken to rest. Age grouping (baby, toddler) is ordinal because the categories have a specific order.
Sampling Methods: Simple Random Sampling [17:15]
The video discusses various sampling methods used to collect data. Simple random sampling ensures that every member of the population has an equal chance of being selected. For example, to select five students from a class of 40, each student must have an equal chance of being picked. This can be achieved by assigning each student a number, placing the numbers in a basket, and randomly drawing five numbers.
Sampling Methods: Stratified, Systematic, and Cluster Sampling [25:04]
Stratified sampling involves dividing the population into subgroups (strata) based on certain characteristics and then taking a random sample from each stratum. For example, a survey in Yuba City could divide the city into four zones (north, south, east, west) and randomly select people from each zone. Systematic sampling involves selecting every kth member from a population after a random start. Cluster sampling involves dividing the population into clusters, randomly selecting some clusters, and then including all members from the selected clusters in the sample.
Sampling Methods: Convenience Sampling and Examples [32:22]
Convenience sampling involves selecting individuals who are easily accessible to the researcher. For example, a researcher studying the effects of drinking and driving might stand outside a bar and ask people questions. The video provides examples to differentiate between sampling methods. Stratified sampling is used when patients are divided by surgery type, and a random sample is taken from each group. Simple random sampling is used when all patients are labeled, and patients are randomly selected. Cluster sampling is used when facilities are selected from seven states, and all patients from those facilities are included. Systematic sampling is used when every hundredth surgery is recorded, and convenience sampling is used when complications from 20 surgeries this week are recorded.
Experimental Design: Explanatory and Response Variables [38:56]
The video introduces the concepts of explanatory and response variables in experimental design. The explanatory variable (independent variable) is the variable that is manipulated or changed to see its effect on another variable. The response variable (dependent variable) is the variable that is measured to see if it is affected by the explanatory variable. For example, if plant food is given to a plant, the plant food is the explanatory variable, and the height of the plant is the response variable.
Experimental vs. Observational Studies [42:00]
The video differentiates between observational studies and designed experiments. In an observational study, data is collected without any attempt to manipulate or influence the variables of interest. In a designed experiment, a treatment is applied to individuals to determine its effect on an outcome. For example, polling students about tuition increases is an observational study, while giving some students a tutor to see if their grades improve is a designed experiment.
Confounding Variables and Types of Observational Studies [46:51]
The video discusses confounding variables (or lurking variables), which affect both variables of interest but are not acknowledged. For example, the amount of sunlight a plant receives can be a confounding variable when studying the effect of fertilizer on plant height. The video also describes types of observational studies: cross-sectional studies (data collected at one point in time), retrospective or case-control studies (data collected from the past), and prospective or cohort studies (data collected in the future).
Types of Experiments: Complete Random Design, Match Pair Design and Randomly Block Design [53:12]
The video describes different types of experiments, including complete random design (each experimental unit is assigned to a treatment randomly), match pair design (experimental units are paired up, and each pair is assigned to a different treatment), and randomly block design (experimental units are divided into groups, and each group is randomly assigned to a treatment). It also discusses the importance of replication, blind studies (individuals do not know which treatment they are getting), and double-blind studies (neither the individuals nor the researchers know which treatment is being administered) to minimize bias.
Potential Biases in Study Design [1:01:13]
The video discusses potential biases in study design, including lurking and confounding variables, overgeneralization (applying results from one group to all groups), cause and effect (assuming one variable causes another just because they are related), sampling error (difference between sample results and true population results), and non-sampling errors (errors from the survey process). It also covers biases in surveys, such as wording of questions, ordering of questions, non-response surveys, and volunteer response.
Examples of Biases in Studies and Surveys [1:10:21]
The video provides examples of biases in studies and surveys. Using different teachers for computer-based homework and traditional homework can bias the results. Assuming a drug safe for one species is safe for all species is overgeneralization. The wording of questions can also bias survey results. The video concludes by emphasizing the importance of careful study design to minimize bias and ensure valid results.