Research and Program Evaluation

T-Test

-A statistical testing method used to determine the probability that, when comparing two separate sample tests with different means, the difference in the means is statistically significant -In other words, researchers can infer that the same difference will be found between the same two groups in the target population as opposed to only being found between the two specific sample sets. -Usually the t-test is only used when the data sets have normal distributions and low standard deviations -The calculated t-test statistic corresponds to a table of probability values -These values indicate the likelihood that the difference between groups is simply due to chance. -Traditionally, if the t-test statistic correspond with less than a 5 percent probability that the differences between the two data sets are by chance, then there's a statistically significant difference between the two sample set.

Accountability

-An important aspect in conducting research -Researchers typically need to have a purpose for conducting their studies -This purpose should be beneficial to the target population, fair, timely, cost-efficient, and should not harm any groups in the process -Funding sources, as well as internal and external auditing groups, may regularly evaluate research operations and outcomes -This is done to ensure that the funds are used appropriately, the research process is ethical, and useful results are obtained -Typically, structured evaluation processes (such as benchmarks and reporting practices) are established and documented during the study's planning and proposal periods

Correlational Designs

-Analyze the strength of the relationship between two variables in one group -One unique type of correlational design is found in ex post facto studies -The researcher examines two existing groups and analyzes the correlation between the variables of interest -Another unique type of correlational design is found in prediction studies, where the researcher determines a correlation between variables, and then uses it to predict other correlations, related events, or future events. -Strength and description of the correlation is indicated by the correlation coefficient (r), which falls between -1 and 1. -If r=0, it indicates there's no relationship between the two variables - r=1 indicates a direct, perfect correlation -If r= a negative value, it indicates an inverse relationship between the two variables. -If r = a positive value, it indicates a direct relationship between the two variables. -Regardless of how strong the correlation is between two variables, it doesn't indicate that one causes the other -Just indicates that these two variables tend to occur (or not occur) together to some degree.

Bivariate Tabular Analysis

-Bivariate tabular analysis is a basic form of analysis used when the value of an independent variables known to predict an exact value for the dependent variable -This is most commonly illustrated by a traditional XY plot graph that marks the independent variable X values across the horizontal axis, and marks dependent variable Y values along the vertical axis. -Once all of the values are plotted, a relationship or lack there of can be seen between the independent and dependent variable

Survey designs

-Can be conducted through paper or electronic questionnaires (either at an external facility or at the study participant's home) -Generally used when research about a particular topic is limited so that more information can be gathered to better shape the research question or topic -Surveys are easy (and usually cost-efficient) to administer, but they can also result in low or biased participant response rates.

Qualitative Research

-Commonly employed in social sciences, including the field of counseling -Typically focuses on the analysis of a group of people (which is sometimes biased) to understand different aspects of human behavior, relationships, and social interactions -Researcher does not manipulate variables when conducting qualitative research -Qualitative research is primarily conducted without rigid structures in place. -Data is collected through the following: -Case studies -Focus groups -Interviews -Observation

Levels of measurement

-Describe the data collected during a study or experiment -Includes: -Nominal -Ordinal -Interval -Ratio

Experimental and quasi-experimental research

-Employs highly controlled processes with the hope of determining a causal relationship between one or more input (independent) variables and one or more outcome (dependent) variables. -Uses random sampling and assignment methods to make inferences for larger populations -Typically, to compares a control group (serving as baseline) to a test group -Ideally, should be able to be replicated numerous times with the same results -Ultimate goal of a well-designed experiment is to declare that a particular variable is responsible for a particular outcome, and that, without that variable, the associated outcome wouldn't occur -Quasi-experimental research employs many of these same qualities, but it often doesn't use random sampling or assignment in its studies or experiments -Consequently, quasi-experimental research produces results that often don't apply to the population at large -They do, however, provide meaningful results for certain subgroups of the population

Inductive Research

-Examines information that's already available (such as establishes data sets like the US census report) to highlight data trends and make inferences and/or projections based on those patterns.

Deductive Research

-Focuses on a specific theory, and then establishes hypotheses to methodically test the theory in order to support or discredit it. -Often involves setting up experiments, trials, or data collection surveys to collect info related to the theory

Systematic random sampling (pros, cons, use when)

-For this type of probability sampling, researchers pick a random integer (n), and then select every nth person from the target population for the research sample -Pros: a simple, cost-effective sampling technique that generally provides a random sample for the researchers. It ensures that sampling occurs evenly throughout an entire target population. -Cons: researchers need to ensure that their original target population (from which the sample is selected) is randomized and that every individual has an equal probability of being selected. Researchers need to be familiar with the demographics of the target population to ensure certain trends don't appear across the selected participants and skew the results. -Use when: a highly controlled experiment setting is necessary; researchers are short on time or funding and need a quick, cost-efficient method to create a random sample

Simple Random Sampling (Pros/cons/use when)

-For this type of probability sampling, the participants are taken directly from a larger population with the characteristics of interest. -Each individual in the larger population has the same chance of being selected for the sample. -Pros: closely represents the target population, thus allowing for results that are in the highest validity -Cons: obtaining the sample can be time consuming -Use when: a highly controlled experiment setting is necessary

Stratified Random Sampling (pros, cons, use when)

-For this type of probability sampling, the researchers first examine the traits of the larger population, which are often demographic or social traits like age, education status, marital status, and household income. -They then divide the population into groups (or strata) based on these traits -Members of the population are only included in one stratum -Researchers then randomly sample across each stratum to create the final sample set for the study -Pros: closely represents the target population, which allows for results that are highest in validity. Since the sampling method is so specific, researchers are able to use smaller samples. -Cons: obtaining the sample can be tedious. Researchers may first need to compile and become acutely knowledgable about the demographic characteristics of the target population before selecting a representative sample -Use when: a highly controlled experiment setting is necessary; demographic, social and/or economic characteristics of the target population are of a special interest in the study; or researchers are studying relationships or interactions between two subsets within the larger population

External Validity

-Illustrates how well inferences from a sample set can predict similar inferences in a larger population (i.e., can results in a lab setting hold true when replicated in the real world). -A sample set with strong external validity allows the researcher to generalize, or, in other words, to make strongly supported assumptions about a larger group -For a sample to have stronger external validity, it needs to have similar characteristics and context to the larger population about which the researcher is hoping to make inferences -A researcher typically wants to generalize three areas: -Population: Can inferences from the sample set hold true to a larger group of people, beyond ht specific people in the sample? -Environment: Can inferences from the sample set hold true in settings beyond the specific one used in the study? -Time: Can inferences from the sample set hold true in any seasonal or temporal period? -If results from the sample set can't hold true across these three areas, the external validity of the study is considered threatened or weak -External validity is strengthened by the number of study replications the researcher is able to successfully complete for multiple settings, groups, and contexts. -External validity can also be strengthened by ensuring the sample set is as randomized as possible

Internal Validity

-Illustrates the integrity of the results obtained from a sample set, and indicates how reliably a specific study was conducted -Strong internal validity allows the researcher to confidently link a specific variable or process of the study to the results or outcomes -The strength of a study's internal validity can be threatened by the presence of many independent variables -This can result in confounding, where it's difficult to pinpoint exactly what is causing the changes in the dependent variables -The internal validity can also be threatened by biases (sampling bias, researcher bias, or participant bias) as well as historical, personal, and/or contextual influences outside the researcher's control (natural disasters, political unrest, participant death, or relocation). -Internal validity can be strengthened by designing highly controlled study or experiment settings that limit these threats.

Type I and Type II Errors

-In experiments, the researcher declares a hypothesis that a relationship doesn't exist between two variables, groups, or tangible instances -This hypothesis is referred to as the null hypothesis -Errors can be made in accepting, or rejecting the null hypothesis based on the outcomes of the experiment. -If the researcher rejects the null hypothesis when it is actually true, this is known as a type I error. -A type I error indicates that a relationship between two variables exists when, in reality, it doesn't. -If the researcher fails to reject the null hypothesis when it's actually false, this is known as a type II error. -A type II error indicates that a relationship between two variables doesn't exist when, in reality, it does. -These errors typically result when the experiment or study has weak internal validity

Sampling

-Is a method of collecting participants for a study. -It's a crucial component of the research design and study process -There are a number of different ways to select samples ,and each method has pros, cons, and situations where it's the most appropriate one to use

Observation

-Researcher swatches the individual or group of interest -A number of additional factors usually shapes the development of the observation study -Researcher can observe participants in a specific situation or highly controlled context, or the researcher can observe the participants in their day-to-day routines. -Participant may or may not know that they are being observed for specific behaviors -Researcher can involve themselves in the contest and become part of the observation study -Researcher can also freely write down data from the observations, or use a pre-made scale or data sheet to document specific behaviors.

Ethical Issues Faced in Doing Research

-Researchers should strive not to harm any population when conducting studies or experiments -All biomedical and behavioral research requires the approval of an independent institutional review board (IRB) to begin -In addition, the IRB monitors the resaserch study for its duration to ensure that populations aren't experiencing any harm -Typically, IRB-approved studies require participant consent form, the guarantee of anonymity and/or confidentiality for all participants and associated data, and strict monitoring and evaluation processes -Researchers should also be mindful of producing, collecting, and publishing accurate, authentic, and non-biased data. -Failure to collect or publish data that does not support the researcher's hypotheses or motives- or that inadvertently brings awareness to another sensitive issue- can also be grounds for ethical concern

Chi Square

-Similar to statistical testing methods like t-tests and ANOVA tests, a chi square test analyzes data between independent groups. -However, chi square tests focus on variables that have categorical data rather than numerical data -They can only be run on data with whole integer tallies or counts, and they're typically used when a researcher has large, normally distributed, and unpaired sample sets

Research

-Systematically investigate an experience either to understand what causes it, to o develop a theory about how that experience can cause a future event. -Systematic investigation can occur through a number of scientific methods -Research designs determine how to structure a study based on factors such as variables being tested, the level to which the designer is manipulating a variable in the study, the types of subjects in the study, what the study is testing or looking for, the frequency and duration of data collection, and whether the data collected is qualitative or quantitative in nature.

Case Studies

-These are detailed and documented examples of the topic of interest -They can be real or hypothetical situations -Often record data over a period of time to examine a specific variable of interest -Can examine a situation involving one individual, a family, a larger community of people, or an organization. -Frequently look at how people relate to one another, and/or to their physical or emotional environment.

Focus Groups

-These bring together a relatively small group of individuals -The group can be diverse in nature or have many similar interests -A facilitator guides a discussion within the group to discern information about individual or collective viewpoints about a specific issue

Analyses of Covariance (ANCOVA)

-This analysis is a type of ANOVA. -This analysis is used to control for potential confounding variables, and is commonly referred to as ANCOVA. -Say a researcher is testing the effects of classical music on elementary students' ability to solve math problems -If the students being tested are in varying grades, then their grade level must be taken into account -This is because math ability generally increases with grade level. -ANCOVA provides a way to measure and remove the skewing effects of grade level in order to better understand the correlation that's being tested

Convenience Sampling (pros/cons/use when)

-This is a type of non probability sampling where researchers select participants who are easily accessible due to factors like location, expense, or volunteer recruitment -Pros: saves times and is cost-effective since researchers can create their sample based on what permits the easiest and fastest recruiting of participants -cons: highly prone to bias. it is difficult to generalize the results for the population at large since the sample selection is not random -use when: conducting initial trials of a new study, when researchers are simply looking for basic information about the larger population (i.e., to create a more detailed hypothesis for future research)

Interviews

-Typically more personal in nature -Can be conducted in person, over the telephone, or via email or regular mail -The interviewer asks individual or group a series of meaningful questions related to the research topic -Can be structured with the interviewer having pre-set questions to ask, or it can be structured with the interviewer asking questions based on the flow of conversation and the answers given by the interviewees

Nonparametric Tests

-Typically used when datasets don't have set parameters, are skewed in distribution, include outliers, or are unconventional in some other way -As a result, nonparametric tests are less likely to be valid in showing strong relationships, similarities, or when differences between groups exist -Its also easier to make a type I error when running nonparametric tests -Some common nonparametric tests include the Mood's Median test, the Kruskal-Wallis test, and the Mann-Whitney test

Post Hoc Tests

-Usually performed after running other tests (e.g., t-tests or ANOVA tests) where it's been determined that statistically significant differences exist between two or more sample sets -At this point, researchers can pick and choose specific groups between which to analyze similarities or differences -Some common post hoc tests are the least significant difference test, Tukey's Test, and confidence interval tests, which are often similar to running multiple t-tests -Post hoc tests can be complex and time consuming to calculate by hand or with simple software, so they often must be completed using sophisticated statistical software packages

Non-Experimental Quantitative Research

-Utilizes logical, empirical methods of collecting information. -This info is called data and is often analyzed using stats -Includes forms of data collection where the researcher collects data that's already available in some form. -Then analyze the dataset to describe the relationship between pre-determined variables. -Researcher does not set up a novel system of trials to produce new data, and they can't randomize any data collected -The researcher has no part in manipulating any variables or establishing a separate control group to which they can compare collected data. -The lack of a control group, lack of variable manipulation by the researcher, and lack of randomization are often seen as weaknesses in non-experimental quantitative research studies -Some examples of non-experimental quantitative research designs: -Survey designs -Correlational designs -Comparative designs

Analysis of Variance (ANOVA)

-Variance tests examine the means of two or more sample sets to detect statistically significant differences in the samples. -Analyses of variance tests (commonly referred to as ANOVA tests) are more efficient and accurate than t-tests when there are more than two sample sets. -There are multiple types of ANOVA tests -One way ANOVA tests are used when there's only one factor of influence across sample sets. -Consequently, two and three way ANOVA tests exists and ae used in the case of additional factors -ANOVA tests can also analyze differences in sample sets where there are multiple dependent and independent variables. -ANOVA tests work by creating ratios or variances between and within the sample sets to determine whether the differences are statistically significant -Calculating these ratios is fairly tedious, and researchers generally use statistical software packages, such as SPSS, SAS, or Minitab to input the datasets and run the calculations -SPSS stands for Statistical Package for the Social Science and is one of the most popular packages that performs complex data manipulation with easy instructions -SAS stands for Statistical Analysis System, and is a software developed for advanced analytics, data management, business intelligence, multivariate analyses, and predictive analytics. -Minitab is an all purpose statistical software created for simple interactive use.

Writing Issues Faced in Doing Research

-When writing research proposals, reports, or manuscripts, researchers may run into issues -These can include inefficient time and project management, incorrect or missing source citations, or failure to acknowledge personnel on final document submissions -When submitting final documents, researchers should be aware of any stylistic, formatting, or content guidelines, as these can vary by agency and publication -It's considered good practice and standard convention not to submit the same manuscript to multiple publications since it's disrespectful to the editors' time. -Also increases the possibility that the manuscript could be published more than once, which could mislead other researchers who may be citing that study -Even with hardwork and due diligence, researchers should be prepared to have final manuscripts returned or rejected -Manuscripts sometimes even require multiple rounds of revisions before publishing body determines that they're ready for publication

Variable

A factor that can be changed

Forms of Hypothesis

A hypothesis typically takes one of two forms: -Null hypothesis: declares there is no relationship between two variables -Alternative hypothesis: declares a specific relationship between two variables, or simply states that the null hypothesis is rejected

Hypothesis

A prediction made in order to lay the foundation for ensuing tests, experiments, or empirical data collection and analysis

Regression Analysis

A statistical technique to determine interactions and relationships between variables, usually to make predictions

Correlation Coefficient

A value between -1.0 and 1.0 that illustrates the strength or interaction between two variables, with zero indicating no relationship.

Independent and dependent variables

A variable is one factor in a study or experiment. An independent variable is controlled by the researcher and usually influences the dependent variable (the factor that's typically measured and recorded by the researcher)

Julia is collecting a stratified random sample to see if level of education correlates with household income in the state of Texas. If 28% of the people in Texas have at least a bachelor's degree, what percentage of Julia's sample set needs to have a bachelor's degree? a. 28% b. 50% c. 78% d. 90%

A: 28% of Julia's sample set needs to have a bachelor's degree. When collecting a stratified random sample, the characteristics of the sample set must exactly match the characteristics of the targeted general population.

Case studies, focus groups, and research observation are all examples of what kind of data collection methods? a. Qualitative data collection methods b. Quantitative data collection methods c. Purposive sampling d. Systematic random sampling

A: Case studies, focus groups, and research observation are all examples of qualitative data collection methods. These methods collect data related to descriptors and details, rather than data related to counts and numbers.

It is good practice and standard convention to submit final manuscripts for scholarly publication to how many journal(s) or publisher(s)? a. One b. Two c. Three d. Four

A: It is good practice to submit final manuscripts for scholarly publication to one journal or publisher. It is important to respect the journal and the editors time, and to prevent the possibility that original research may be published more than once. This can cause major reference and citation errors for future researchers, especially those conducting meta-analyses.

Joe is doing a research study on the effects alcohol consumption has on long-term mood. His control sample self-reports not drinking alcohol. His test sample self-reports drinking at least one glass of wine every night. He collects self-reported data on each group's mood for 60 days. Both data sets have normal distribution and low standard deviation. Joe now wants to test the statistical significance of the difference in mood between the control group and the test group. Which test should Joe conduct? a. T-test b. Chi-square test c. ANOVA test d. Post hoc test

A: Joe should conduct a t-test. A t-test examines differences between two (and only two) normally distributed groups with low standard deviation to determine if the differences are statistically significant, or due to chance.

A study is considered to have strong external validity if the researcher can replicate and generalize its findings across multiple instances of what three factors? a. Population, time, and environment b. Gender, age, and height c. Animal studies, human studies, and non-living object studies d. Infant years, adolescent years and adult years

A: Population, time, and environment. If a research design can produce the same outcomes across multiple types f people, multiple seasons or temporal periods, and across a variety of settings and situational contexts, it is considered to have high external validity. Any results are likely generalizable to the target population.

A type I error __________ a ____________ null hypothesis a. Rejects; true b. Rejects; false c. Fails to reject; true d. Fails to reject; false

A: Rejects; true. A type I error in statistical testing occurs when the researcher rejects a null hypothesis that is, in fact, true. This type of error in a test will indicate that a relationship occurs between two variables when it actually doesn't.

A researcher notes changes in the outdoor temperature over a one-month period. The recorded temperatures are an example of what type of measure? a. Interval b. Nominal c. Ordinal d. Ratio

A: The recorded temperatures are an example of an interval measure. An interval measure describes variables that fall at equally spaced intervals.

Nonparametric tests are more likely to result in which of the following? a. Type I error b. Type II error c. Confounding d. Independent variables

A: Type I error. Since nonparametric tests work with data sets that tend to be noisy in some way (outliers, abnormal distribution, etc.) they may show an effect between variables when there isn't actually one.

Confounding

An effect that occurs in an experiment or study due to the presence of an extra variable that can influence or correlate with the established independent or dependent variables

Purposive sampling (pros/cons/use when)

Another non-probability sampling method used when researchers have a precise purpose or target population in mind. -Pros: helps increase recruitment numbers in otherwise hard to access populations -Cons: usually unable to generalize the results to larger populations beyond the sample's specific subset -Use when: researchers have a precise purpose for the study, or a specific group of participants is required that isn't easy to select through probability sampling methods

The acronym ERIC stands for what? a. Electronic Research Institute Collaboration b. Education Resources Information Center c. Education Research Information Center d. Electronic Research Incorporated

B: ERIC stands for Education Resources Information Center. ERIC is a scholarly online database funded by the United States Department of Education and contains over 1.5 million pieces of high-level educational literature dating back to 1966.

A researcher is grouping participants into sample sets by gender. In this instance, gender is an example of which type of measure. a. Interval b. Nominal c. Ordinal d. Ratio

B: In this case, gender is an example of a nominal measure. A nominal measure describes variables that are categorical in nature.

One fundamental difference between probability sampling and non-probability sampling is that probability sampling uses ___________ while non-probability sampling does not. a. Convenience b. Randomization c. Volunteerism d. Referral

B: Probability sampling uses randomization. In probability sampling, anyone who is eligible to be in the sample has an equal chance of getting randomly selected, whereas non-probability sampling may choose participants based on convenience, referral, volunteerism, etc.

Rachel is going out for lunch. As she leaves her office, she wonders if eating veggie tacos for lunch has any effect on how many data sets someone can analyze in a given afternoon. When she returns from lunch, she conducts a literature review to see what data is available on the topic. She finds none. Over the next few weeks, when any of her colleagues leave for the day at the same time as she does, Rachel asks them what they ate for lunch and how many data sets they analyzed after 1:00pm. The type of sample Rachel is collecting is referred to as which of the following? a. Simple random sample b. Convenience sample c. Systematic sample d. Stratified random sample

B: The type of sample Rachel is collecting is referred to as a convenience sample. Rachel is collecting information from people who work with her and who sees at a good time, so she is collecting samples based on proximity. This sample is not indicative of the general population and can be biased, but it could provide Rachel with enough information for more rigorous testing in the future.

Tina is conducting a research study about underage drinking habits on college campuses on football game days. As she's scoring her collected data, she notices the test group is primarily made up of freshman students and the control group is primarily made up of sophomore students. What does Tina realize about this distinction? a. It's to be expected b. It may confound her results c. It won't make a difference if she runs enough ANOVA tests d. It should't be tabulated in a chi-square cross section

B: Tina realizes that this distinction may confound her results. Confounding occurs when any results found correlate with multiple variables, therefore preventing a clear correlation between a particular independent variable and a particular dependent variable. In this instance, the fact that a majority of the test group and control group are made up of a particular year of students may influence the presence of any habits that Tina is trying to study.

Empirical

Based on observation, experience, and/or experimentation

Joanna is interested in the attitudes, beliefs, and perceptions of working fathers towards the quality and affordability of childcare. She books a conference room in her office, invites 15 fathers working in her building to attend a two-hour session, and presents them with a series of questions to discuss with each other. They can also direct any questions and comments about the topic to Joanna. She mainly documents the conversation, but if there's a lull in the conversation she joins in to facilitate the discussion. This is an example of conducting which of the following? a. Case study b. Survey c. Focus group d. Observation

C: A focus group hopes to facilitate discussion of a particular topic between persons of interest, and the researcher typically collects relevant notes from this discussion.

Anna is doing a research study on the effect various study habits have on test scores. She collects four sample sets from a Psych 101 class of 400 hundred students and has each group practice a different habit while studying for the first exam. The first sample group listens to classical music during every study session. The second sample group exercises for 30 minutes before every study session. The third sample group naps after every study session. The fourth sample group drinks coffee before every study session. Anna collects exam scores for every participant and wants to test the statistical significance of the differences between scores for the sample groups. Anna should conduct which of the following tests? a. T-test b. Chi-square test c. ANOVA test d. Post hoc test

C: Anna should conduct an ANOVA test. An analysis of Variance test is the most efficient and most accurate way to test for statistically significant differences between three or more sample sets.

All biomedical and behavioral research proposals require approval by whom? a. A government employee b. A six-person committee of licensed research physicians (typically those in the field of the research topic) and/or psychologists c. An institutional review board d. A participant from the sample group.

C: Biomedical and behavioral research proposals require approval by an institutional review board. An institutional review board is typically an independent committee focused on the ethics of the research proposal. This board approves the initial proposal, then monitors the project for its duration to ensure no harm comes to research subjects.

Which of the following is true given the advent of technology in the counseling field? a. Technology has advanced the counseling field because it eliminates geographical barriers and can be less tedious to document cases b. Technology has devastated the counseling field by making counseling sessions more robotic and by lacking the personal touch of an in-person session c. Currently, there is not sufficient evidence as to whether or not technology is advantageous or disadvantageous to the counseling field d. Technology is seen as a necessity to the counseling field, as it allows easier access to clients.

C: Currently, there is not sufficient evidence as to whether or not technology is advantageous or disadvantageous to the counseling field.

Bill is conducting a research study on workplace views of maternity leave in a specific industry. His funding source requires that at least 60 percent his sample set be women. The industry that he's studying consists of 32 percent women, and he has to search a little harder to find enough female participants. This is an example of what type of sampling? a. Simple random sampling b. Stratified random sampling c. Ad hoc sampling d. Systematic random sampling

C: In ad hoc sampling, the sampling cannot be randomized if the researcher needs to meet a certain quota for whatever reason. This makes it easier to get the necessary participants, but will lower generalizability for the target population.

Quasi-experimental research methods follow most processes associated with experimental research methods except for ______________; therefore; any results obtained usually cannot be ______________. a. Data synthesis; analyzed b. Trial and error practices; used c. Randomization; generalized d. Clinical trial studies; medical

C: Randomization; generalized. Though quasi-experimental research methods usually don't produce results that can be applied to larger populations, they often produce results that can be applied to larger populations, they often produce results that can be applied to a specific subset.

Who funds ERIC? a. A committee of universities and colleges form across the United States b. Google c. The United States Department of Education d. The United States Congress

C: The United States Department of Education funds ERIC. The DOE has been sponsoring ERIC since the 1960s.

A physician administers a new pain medication as part of a clinical trial, and later asks the participants to rank their pain on a scale of 1 to 5. The patients' responses are an example of what type of measure? a. Interval b. Nominal c. Ordinal d. Ratio

C: The patients' responses are an example of an ordinal measure. An ordinal measure describes variables that fall on a scale or can be ranked.

The correlation coefficient value falls between which values? a. Zero and 100 b. -5 and 5 c. Zero and 1 d. -1 and 1

D: -1 and 1. The correlation coefficient value is always between these two numbers, with zero representing no correlation. A negative value indicates an inverse correlation, and a positive value indicates a direct correlation.

Which of the following is a popular statistical software program? a. SPSS b. SAS c. Minitab d. All of the above

D: All of the above. SPSS, SAS, and Minitab are all popular statistical software programs. These programs make it easier to analyze and run tests on large datasets.

Which visual representation is most often used to present bivariate tabular analyses? a. Parabola b. Pie chart c. Cross-sectional 2x2 table d. XY graph

D: An XY graph is most often used to present bivariate tabular analyses. Bivariate tabular analyses are typically depicted this way since they calculate data points representing independent variables (plotted on the horizontal X-axis) to predict data points representing dependent variables (plotted on the vertical Y-axis)

A type II error __________ a ____________ null hypothesis. a. Rejects; true b. Rejects; false c. Fails to reject; true d. Fails to reject; false

D: Fails to reject; false. A type II error in statistical testing occurs when the researcher fails to reject a null hypothesis that is false. This type of error in a test will indicate that there is no relationship between two variables when there actually is.

A researcher is analyzing trends in insurance reimbursement for counseling costs from 1990 to 2010. The changes in reimbursement are an example of what type of measure? a. Interval b. Nominal c. Ordinal d. Ratio

D: The changes in reimbursement are an example of a ratio measure. A ratio measure describes variables that can have a true zero point (zero dollars and zero cents).

Quantitative

Focusing on the numerical, mathematical, statistical, or otherwise quantifiable aspect of a phenomenon

Qualitative

Focusing on the qualities and/or descriptors of a phenomenon

Ad Hoc Sample (pros/cons/use when)

For this type of non-probability sampling, researchers must meet a set quota for a certain characteristic and can recruit any participant as long as they have the desired characteristics -Pros: allows for greater inclusion of a population that might not otherwise be represented -Cons: results won't be indicative of the actual population in an area -Use when: it's necessary that a group within the larger population needs a set level of representation within the study

Dependent Variable

In an experiment or study, the variable that changes due to manipulation of the independent variable.

Control Group

In an experiment or study, this group mirrors the qualities of the test group as much as possible, but does not receive the test or intervention in order to serve as a comparison

Probability Sampling

Sampling method that chooses sample sets at random; generally more likely to represent the target population

Non-probability sampling

Sampling method that doesn't choose sets at random

Education Resources Information Center (ERIC)

The Education Resources Information Center (ERIC) is a scholarly online database funded by the United States Department of Education. It contains over 1.5 million pieces of high-level, educational literature dating back to 1966.

Mean

The central value of data distributon

Statistically significant

The probability of a relationship, result, or event occurring due to something other than chance

Correlation

The relationship or degree of interaction between two variables

Comparative designs

These examine data trends determine a relationship in two groups or datasets that have already been established

Ratio

This measurement describes anything that has a true "zero" point available (e.g., angles, dollars, cents)

Nominal

This measurement describes variables that are categories (e.g., gender, dominant hand, height)

Ordinal

This measurement describes variables that can be ranked (e.g., Likert scales, 1 to 10 rating scales)

Interval

This measurement describes variables that use equally spaced intervals (e.g., number of minutes, temperature)

Observation

Watching and documenting an experience in detail

Theory

Widely accepted principle or set of ideas that explain an event or events (though a theory can be disproven)