Reading...
Play button
Play button
Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
83 Cards in this Set
- Front
- Back
|
Qualitative research:
|
This method emerged about a century ago in sort of a response to "quantitative" research. The idea today is that asking people simple yes or no questions or numeric questions as they do with Likert scales or most survey questions, don't often provide detailed answers about the why and how of people's lives and decisions.
Theoretically, qualitative research includes spending more time with the subject. Letting the subject tell you how they feel. It's fundamentally about listening. You record and transcribe the results and place your findings into categories. Those categories emerge over time in a grounded way. Qualitative research is extremely common in media research, helping writers and marketers craft messages. Without getting into the logic, qualitative research in recent decades has begun to be less objective in some forms (not in our class, however), taking a stand to be an advocate for the voiceless or downtrodden as the researcher sees it. The method can become a sort of advocacy or activism. |
|
|
Quantitative Research
|
The idea of quantitative research comes from the word "quantity." You are trying to quantify things. You want to "quantify" what a certain population thinks, feels or does. So, when you take a random survey of, say 300, BYU-Idaho students, you are using them to represent -- "to quantify" what typical, average BYU-Idaho students wish.
Quantitative research is "generalizable" to a large population. Quantitative research is excellent for determining simple answers -- with numbers. Quantitative research involves statistics -- descriptive and inferential to show relationships, averages and correlations. Types of quantitative research: Data mining, content analysis and survey are most common for this class. Case studies may include quantitative research. Experiments by their nature are quantitative because we mathematically compare results. In applied research -- Nielsen ratings, QR scores, comScore ratings, index numbers, Kwant scores and similar data are all quantitative by nature. |
|
|
Action Research
|
Rory O.Brien at the University of Toronto gives this straightforward definition of action research:
"Put simply, action research is 'learning by doing' - a group of people identify a problem, do something to resolve it, see how successful their efforts were, and if not satisfied, try again." (Citation: O'Brien, R., "An Overview of the Methodological Approach of Action Research, In Roberto Richardson (Ed.), Teoria e Prática da Pesquisa Ação Theory and Practice of Action Research. João Pessoa, Brazil: Universidade Federal da Paraíba. (English version) Available: http://www.web.ca/~robrien/papers/arfinal.html and at http://www.web.net/~robrien/papers/arfinal.html, accessed Jan. 11, 2012). Action research involves those involved in a problem working to solve it -- measuring results and trying again. It differs from other research in that most research the researcher is an outsider. Also, action research differs from normal problem solving in that it tries to systematize and follow general research methods, ra |
|
|
Basic research
|
Basic research is research to answer questions about which we are curious. It is the kind of research that leads to creation of encyclopedias. It answers important questions of science, builds on existing knowledge. It is the type of research that gets published in academic journals. The specific use of that research is unclear, at least at first. It is, likely, the type of research you think about when you think about scientific research.
|
|
|
Applied research
|
Applied research is research designed to answer specific questions or solve problems. (Unlike action research, applied research is done by outsiders in a more-or-less objective manner.) Applied research in communication is all market research about who watches shows or what advertising drives traffic to websites or questions like that. It usually comes from an organization's need rather than general curiosity. Professionally, applied research is rarely printed in an academic journal while applied research is what most communication professionals do.
The Edsel case, the Walmart case, the LineTamer case were all examples of applied research, or lack thereof, designed to answer specific questions needed by an organization or a society. While the research methods and techniques are the same between applied and basic research, the purposes and the way such research is published marks the main differences between them. |
|
|
Meta Question
|
The Meta-question is the "big" over-arching, broad question you want to know more about. It is the first step of many research projects. A useful meta-question can't be answered with one research project, and may, indeed, never be fully answered?
This idea of meta-question is a concept that not every text or scholar uses, but I find it helps all students understand how to start research. Here is an example: How do violent video games affect relationships? In our example, the concept of a meta-question about violent video games and relationships would lead to researching encyclopedias, books and academic papers to see what has been written and studied about that important question. Your research provides a partial answer to that meta-question. With that partial answer, you can then create a little study of that you can do to enhance the world's and your own understanding of this big question you have. |
|
|
Research questions
|
Research questions (and hypotheses, which follow in the next entry) are the specific things you wish to answer in a scientific project.
Unlike a "meta-question" which is broad, a research question is specific enough to provide a narrow, concrete answer answer. In our example of violent video games, let's say you craft a complex experiment where you watch people interact in a tense situation after they've played a violent video game -- with a control group who didn't play the game. Let's say you are looking to see whether someone yells or not in the tense situation. Here is a research question: Are people more likely to yell in a tense situation after playing a violent videogame? Such a question is specific and can provide a concrete answer. Most research questions in qualitative research include a how or why in them. |
|
|
Hypotheses
|
A hypothesis is a prediction. It is what you think will happen. Your experiment is how you confirm whether the hypothesis holds up -- or is significant. (I purposely avoided the term "is true" because science never works that way. You only become mostly sure something is true.)
In our research question about violent video games and tense situations, you'd make a hypothesis as a statement similar to this: H1 (That is Hypothesis one): Those who play violent video games before the tense situation are more likely to yell than those who don't. You will test you hypothesis -- using a mathematical test called either a Chi Square test or a T Test to demonstrate your results. Comparing hypotheses and research questions: a. Hypotheses are by nature involving the relationships between things. They are quantitative and relational. Research questions may be either qualitative or quantitative, applicable for any research project. Research questions tend to be associated with qualitative study, however, or descripti |
|
|
Theoretical
|
Social research is theoretical, meaning that much of it is concerned with developing, exploring or testing the theories or ideas that social researchers have about how the world operates.
|
|
|
Empirical
|
based on observations and measurements of reality -- on what we perceive of the world around us.
|
|
|
Nomothetic
|
Nomothetic refers to laws or rules that pertain to the general case (nomos in Greek) and is contrasted with the term "idiographic" which refers to laws or rules that relate to individuals (idios means 'self' or 'characteristic of an individual ' in Greek). ..."most social research is concerned with the nomothetic the general case rather than the individual. We often study individuals, but usually we are interested in generalizing to more than just the individual.
|
|
|
Probabilistic
|
Probalistic means based on probabilities. The inferences that we make in social research have probabilities associated with them -- they are seldom meant to be considered covering laws that pertain to all cases. Part of the reason ... statistics become so dominant in social research is that it allows us to estimate probabilities for the situations we study.
|
|
|
Causal
|
The term causal means that most social research is interested (at some point) in looking at cause-effect relationships. This doesn't mean that most studies actually study cause-effect relationships. There are some studies that simply observe -- for instance, surveys that seek to describe the percent of people holding a particular opinion.
|
|
|
Variable
|
A variable is any entity that can take on different values, according to our text. That need not be a number (although numbers can be assigned for mathematics later.) For example, The variable: Status in school can vary between Freshman, sophomore, junior and senior. I might then assign 1 to Freshman, 2 to sophomore, 3 to junior and 4 to senior.
|
|
|
Attribute
|
An attribute is a specific value a variable.
It is one of the set of things that the variable can be. In the case of school status, the variable is school status. The four individual attributes of the variable school status are freshman, sophomore, junior and senior. As our textbook says, the attributes of a variable should be exhaustive, no other possibilities and mutually exclusive, meaning they don't overlap. Is it possible that a person could be both a junior and a senior? Not really. |
|
|
Independent variable
|
is a concept only when you are comparing things.
According to our text, the independent variable is what you (or nature) manipulates -- The "cause" if you will. This may seem confusing, but it is the thing you manipulate as a researcher too. (Read more on this after the entry on dependent variable.) |
|
|
Dependent variable
|
The dependent variable is what is affected by the independent variable -- your effects or outcomes.
Say you had two variables: School status and scores on my exams. If you wanted to see if someone later in school does better on the exam, you are doing research. As you look at year in school (the variable you are manipulating, so the independent variable), you are looking to see how the scores change the dependent variable the result. |
|
|
Correlation
|
From our text: "Correlational relationship says that two things perform in a synchronized manner. For instance, there has often been talk of a relationship between ability in math and proficiency in music."
A correlation is when one things goes up, the other goes up (or down) to some measurable degree. The relationship may not be total, but it is real. From the text example. As participation in music rises, math scores rise. |
|
|
Causal relationship
|
One causes the other.
Causes are rare, and rarely studied, in our type of research findings. Avoid thinking that way. Relationships suggest a piece of a puzzle. Multiple causes sometimes are hard to sort out. Sometimes cause is the wrong way to think about it. Say scores on my test grew as years went by. Is that you went from being a freshman to a sophomore really the "cause" of that change? |
|
|
Third variable problem
|
The third variable problem is when there is a relationship between two variables but a third, unknown variable is causing both the independent and dependent variable to rise or to change. We always have to be aware of this problem and, therefore, not assume a "causal" relationship when you only know of a "correlation" for sure.
|
|
|
Positive relationship
|
This is when high values of one variable (or maybe better said higher attributes) go with high values (or attributes) with another variable.
Put another way, when the value of the independent variable rises and the rise correlates with a rise in the dependent variable, you have a positive relationship. |
|
|
An inverse relationship
|
An inverse relationship, often called a negative relationship, is when the value (or attributes) of one variable rise, the value (or attributes) of another variable decreases.
Put another way, when the value of the independent variable rises, the value of the dependent variable declines, you have an inverse, or negative, relationship. An inverse relationship might be when you spend more time playing world of warcraft, your grades go down. Independent variable here is the amount of time spent playing WOW while the dependent variable is GPA. |
|
|
A cross-sectional study
|
is one that takes place at one in time. To quote our text, "In effect, we are taking a 'slice' or cross-section of whatever it is we're observing or measuring. "
|
|
|
Longitudinal study
|
"A longitudinal study takes place over time with more than one slice of information."
An example of a longitudinal study might be when you look at how a person's opinions change over time. |
|
|
Descriptive study
|
A study that merely measures frequencies of things or describes what is there.
My dissertation was largely a descriptive study. It showed how frequently certain attributes of Mormonism were mentioned during the 2008 election involving Mitt Romney. I did limited other analysis. One finding -- that polygamy was mentioned in about one in four stories, was an interesting finding, one that supported a conclusion that reporters often distorted our faith in the coverage. It might be fair to say that most qualitative studies are descriptive by nature. |
|
|
Relational study
|
A relational study looks at the relationship between more than one variable. Basically, this is when we put in a dependent and independent variable and analyze them.
In my doctoral research, I found a relationship between certain ways of describing Mormonism and how favorable those I had read those articles thought Mormonism was portrayed. It seems common sense, but polygamy was inversely correlated with how favorably my researchers thought Mormonism was portrayed. |
|
|
Causal research
|
This kind of research often follows earlier descriptive and relational research. Its point is to show that one variable causes another. It tries to eliminate the third variable problem as much as possible. Experimental methods are common here.
Even so, we may still use the term correlation to describe our findings. An example of a causal study is one I sometimes mention in class where a set of researchers divided a group of students into one of two groups. One group played a violent video game and later answered a questionnaire about research assistants. A second group didn't play the video game. They found those playing violent video games were more likely to be aggressive and judgmental than those not playing the games. Even though researchers never said it "caused" more aggression, the implications were clear enough because the experiment removed many problems with the study methodology. |
|
|
Research proposal
|
The Research proposal is a paper you put together that describes how you will do a research paper, the methods you will use, the samples you will pick, the size of population and how you will answer the question. Most research proposals include a short analysis of what the literature says, what kinds of studies have already been done.
|
|
|
Correlation coefficient
|
This is a mathematical number -- expressed usually as and r or and R-squared value. It shows the strength of a relationship. In an earlier example, we talked about how as one thing rose, another thing rises or falls. The coefficient shows how much of the rise of one is shown in the other. A 1 means that one perfectly predicts the rise of the other. A zero means that one produces no effect at all in the other. A negative 1 means that a rise in one perfectly predicts a fall in the other. Coefficents can't be higher than 1 or lower than - 1.
|
|
|
Population
|
The group of people or items being studied in a quantitative study. So, a population might be BYU-Idaho students or BYU-Idaho women or married Americans or American adults or adults making more than 35,000 a year or unemployed Idahoans or victims of a given disease or corporations or organizations or new articles of a certain type. ... any group that a researcher wished to understand.
|
|
|
Sample
|
A sample is a portion of a population that a researcher studies. They uses samples to make inferences about the population at large.
|
|
|
A Random sample
|
A random sample is where every member of a given population has an equal chance of being selected. Random may be among the most misused and mis-understood terms in science. It does NOT mean haphazard or arbitrary or "coming out of nowhere." It has a specific scientific meaning -- people have an equal chance of being selected.
Excel and the website random.org are excellent ways of developing random numbers for sampling. |
|
|
Non-random Sample
|
A non-random sample is one that is selected for other reasons, but isn't random. These kind of samples are limited in how much they represent a larger population, that is how representative and generalizable they are. They still provide powerful insight.
|
|
|
Purposive sample
|
A purposive sample is one where you select a group of people on purpose. Sometimes you use a purposive sample because you can't get a random sample of a given population say people using drugs. Instead, you try to find a group that has characteristics of those using drugs -- say those convicted of possessing drugs. A purposive sample is deliberate and does the best a researcher can.
|
|
|
Stratified sample
|
A stratified sample takes the population and divides it into groups like education status or gender or grade or something like that, important to the research. Then it takes a random sample of those groups. If 39 percent of our school is freshmen and I did a stratified sample, I would make sure than 39 percent of the people in my study's sample are freshmen. See: The Psychology Department of the University of California, Davis. http://psychology.ucdavis.edu/sommerb/sommerdemo/sampling/types.htm
|
|
|
Quota sample
|
A quota deliberately includes a certain portion, a "quota" of a subsection of a population because I want to include that groups views in my study. That quota may or MAY NOT represent their proportion in the population. For example, let's say you wanted to understand racial attitudes at a campus like BYU-Idaho. If you wished to see the differences between people of color and white students, you might use a quota sample because white students so predominate on this campus. Quota samples require careful reporting in a paper to let an audience know that the results aren't necessarily representative of the population.
See: The Psychology Department of the University of California, Davis. http://psychology.ucdavis.edu/sommerb/sommerdemo/sampling/types.htm |
|
|
Snowball sample
|
A snowball sample is where a researcher asks study participants/study subjects to find others who might be will to participate -- such as their friends or acquaintances. This kind of sampling is useful when studying hard-to-find populations like, say, troubled segments of the poor or homeless.
|
|
|
Convenience sample
|
A convenience sample is the worst of all samples for generalizability. You take who is available or convenient and study those people. It might be a group of friends or people sitting in a class or a group of dorm associates. Sometimes, however, you can't get a group of people any other way, so convenience sampling becomes a default.
|
|
|
Bias
|
Bias is a problem with a sample or a study that may weaken or invalidate the results of a study. Bias is usually not deliberate nor does it reference to the deliberate insertion of personal opinions as it does in the everyday term political or media bias. Like random, bias is a frequently misunderstood word in science.
|
|
|
Selection bias
|
is one of the most common forms of bias. In surveys or studies, selection bias refers to people who choose to take a study for reasons of their own. Selection bias can make people less likely to answer a study or more likely. A good example of selection bias occurs on ESPN's online polls. They will ask which team will win the Super Bowl, or something like that. ESPN breaks it down by state and region and you can see that states where teams are located will more likely vote for their favored team. There may be a selection bias at play in those areas.
|
|
|
Frequency Distribution
|
The most common type of quantitative research is frequency distribution which is really another word for counting. In a survey, I find 29 sophomores, 10 juniors and 5 seniors, for example. That is the frequency distribution of class in school. Let's say I ask a 5-part Likert scale question about the perceptions of a movie. The frequency distribution is merely the number of 1s, 2s, 3s, 4s and 5s.
Just because frequency distribution is simple doesn't make it powerful. For example, from those frequencies, I can argue the degree to which people liked a movie. And even more complex relational or even causal studies can include frequency data as part of their findings and results. |
|
|
Operational Definition
|
The idea of an operational definition is that you find a way to define a concept in such a way that you can research it. Let's say you are studying depression. To study a treatment for depression you must have a way of defining depression such that you can see if the treatment helps. Maybe an operational definition would be a simple self-reported scale saying how frequently a person says they depressed. It's a separate issue whether that is a good definition. It is a definition that you use to say a person "is depressed." That is the idea of operational definition.
|
|
|
Reliability
|
The idea of reliability is that is you did a study the same way, you would get the same results.
|
|
|
Validity
|
Validity is the extent to which a study measures the real world. It may be reliable, that is to say repeatable, but not truly measure the real world.
|
|
|
Self-serving Bias
|
Let's quote wikipedia here: "A self-serving bias, sometimes called a self-serving attributional bias, refers to individuals attributing their successes to internal or personal factors but attributing their failures to external or situational factors. This bias is a mechanism for individuals to protect or enhance their own self-esteem. For example, a student who attributes a good grade on an exam to his or her own intelligence and hours of studying but a poor grade to the professor’s poor teaching ability and unfair test questions is exhibiting the self-serving bias." That bias is important in looking at cause an effect in case studies and depth interviews. When a person is at fault, they are less likely to blame themselves.
|
|
|
Inferential Statistics
|
Inferential statistics are mathematical equations and concepts that you test on your sample to imply and understand your population. Means. Correlations. Standard deviations. Regression. These are examples of inferential statistics.
|
|
|
Level of Significance
|
Level of significance is a statistical term that means how likely it is for something to be random. In short, a high significance level means it isn't likely that something is random. A .95 significance level means that there is a 95 percent chance something isn't random.
|
|
|
p value
|
A p value is the opposite, really, of the level of significance. It shows the liklihood that something would be random. A p value is the answer to the T test in excel. (Don't confuse a p value with a correlational coefficient.) A p value of lower than .05 means there is a 5 percent chance of being random. IT is statistically significant. A p value is expressed as a decimal.
|
|
|
Mean
|
Mean means average. Add up the total numbers and divide by the number of entries in a series. A mean is the most common statistic.
|
|
|
Median
|
A median means middle. If you list all numbers in a series from smallest to greatest, it is the number exactly in the middle of the list. If there is an even number of numbers in the list, take the two middle items and add them. Then divide that sum by two. That is the median. Means can be susceptible to unusually large outliers, especially in small populations. For example, if Bill Gates walked into a room of college students, the mean net worth would be a billion dollars or more per person. A median would be more "typical," a more correct assessment of what the demographics of that room really looks like. So medians are important measures.
|
|
|
Mode
|
The mode means the most frequently occurring number in a series of numbers.
|
|
|
Standard Deviation
|
Standard deviation is a concept describing how close a given sample mean might be to the actual mean one mean might be. It is a measure of what is unusually small and unusually large. Put yet another way, standard deviation is a measure of how far apart numbers are from one another. A small standard deviation means numbers in a study are closely bunched. A large std dev means they are spread out. One standard deviation larger and one standard deviation smaller than the actual mean comprise 68 percent of the total answers. If a number is more than two standard deviations away from the mean, then it is considered statistically significant.
For more information, see this excellent entry on the Internet: http://www.mathsisfun.com/data/standard-deviation.html |
|
|
Statistically Significant
|
This, like random and bias, has a meaning a little different from the meaning we might use in everyday conversation. Significant here means that it is unlikely to be random. The p value is less than .05, so it is significant. The level of significance is greater then .95, so it is statistically significant. Put another way, statistical significance is where you are saying "something seems to be going on here. It isn't just something that happens at random. This might be a cause of something. It correlates with something." Put one more way: When something is statistically significant, as the independent variable changes, so does the dependent variable in ways that can be predicted.
|
|
|
Type I Error
|
A type one error occurs when a researcher says a hypothesis is false when it is actually true
|
|
|
Type II Error
|
A type two error occurs when a researcher says that a hypothesis is supported and seems true, but it isn't. When a researcher has a significance level .of .95 and believes a hypothesis to be supporter, there is still a .05 percent chance of error, a Type 2 error. More detailed statistics than provided in this class look to cut Type 1 and Type 2 errors through something called Analysis of Variance or ANOVA.
|
|
|
Control Group
|
A Control group is part of an experiment. It is the group that does not receive a stimulus -- the experiment. For example, let's imagine that you are trying to show in an experiment how violent video games affect how well a student does on a test. One group would play the games and the take the exam. The other would not play the games but still take the exam. the second group, the one that doesn't take the exam, is the control group. By comparing the means of the scores, you can see whether their appears to be an effect or correlation.
|
|
|
Factor
|
For our purposes, factor is a synonym for variable.
|
|
|
t test
|
A T test is a statistical test to see the likelihood that a given mean is random compared with the actual mean. (It is effectively a mean.) In Excel, you calculate a T test as =ttest and then select two sets of numbers that have the same range. When the result of a Ttest is .05 of less, then the test is statistically significant. The math behind a Ttest is complex, but you don't need to know the math, just =ttest in Excel.
|
|
|
Regression to the Mean
|
Regression to the mean is a mathematical concept that implies that outlying samples will "regress to the mean" if selected again. For example, if you are unusually tall, chances are your children, while taller than average will mostly be closer to the mean than you are. Here is wikipedia's solid definition: "In statistics, regression toward the mean (also known as regression to the mean) is the phenomenon that if a variable is extreme on its first measurement, it will tend to be closer to the average on a second measurement, and—a fact that may superficially seem paradoxical—if it is extreme on a second measurement, will tend to have been closer to the average on the first measurement.
|
|
|
Replication
|
Replication is qualitative research's replacement for reliability. If two studies looking at the same research topic achieve similar results, they are considered replicable.
|
|
|
Authenticity
|
Authenticity is the degree to which findings in a qualitative study were presented with accuracy. In a sense, it is qualitative analysis's similar term to validity in quantitative analysis. Here is a definition from Medscape.com: "Authenticity focuses on the degree to which researchers faithfully and fairly described participants' experiences." (See: http://www.medscape.com/viewarticle/712876.)
|
|
|
Emergence
|
In qualitative analysis, emergence is the process by which themes and concepts arise. As you analyze, think about and sort core concepts in a qualitative study, you state what emerged from those data and from your analysis.
|
|
|
Grounded Theory
|
Wikipedia's definition of Grounded theory is excellent: "Grounded theory method (GT) is a systematic methodology in the social sciences involving the discovery of theory through the analysis of data. It is mainly used in qualitative research, but is also applicable to quantitative data.
Grounded theory method is a research method which operates almost in a reverse fashion from traditional social science research. Rather than beginning with a hypothesis, the first step is data collection, through a variety of methods. From the data collected, the key points are marked with a series of codes, which are extracted from the text. The codes are grouped into similar concepts in order to make them more workable. From these concepts, categories are formed, which are the basis for the creation of a theory, or a reverse engineered hypothesis. This contradicts the traditional model of research, where the researcher chooses a theoretical framework, and only then applies this model to the phenomenon to be studied." |
|
|
Interviewing
|
Interviewing is a qualitative technique where participants receive the same general questions, one by one. Their answers are recorded and answers given.
|
|
|
Case Study
|
Case study is generally considered a qualitative research method. A researcher looks at a situation/experience and tries to provide insight into why something occurred the way it did, as a way to explore reasons for success of failure. A case study uses analysis of data, depth interviews and insightful research questions. It is usually written as a narrative. Many, though not all, case studies conclude with an analysis and set of takeaways. Some just write a narrative and leave it there.
|
|
|
Ethnography
|
Here is the dictionary.com definition: "a branch of anthropology dealing with the scientific description of individual cultures." In communication research, ethnography looks at how media shapes culture, including popular culture. Ethnography is usually qualitative. A person doing ethnographic research might look at how texting shapes lives.
|
|
|
Narrative Analysis
|
Narrative, or textual, analysis is a qualitative method we likely won't use in this class. However, it is growing in popularity. A narrative analysis looks at how information is presented and looks at that information as a story with plot, hero and villain. Looking at it that way, we can see values, concepts and relationships. Successful narrative analysis requires a philosophical point of view from which the analysis starts, which is why we likely don't do it at this level. Such analysis is a great way of thinking about the thinking behind the information is presented. Some consider narrative analysis as a type of case study. A very worthwhile powerpoint on this subject, if you wish to know more, is online at: http://www.methods.manchester.ac.uk/events/whatis/narrativeanalysis.pdf
|
|
|
Phenomenology
|
Phenomenology is a big word that basically means the study of people's perceptions of phenomena. It is how people interpret the world around them. It is a qualitative approach. In general, as you do depth interviews or focus groups you may well be doing a type of phenomenology.
|
|
|
One-shot design
|
One shot design is another word for cross sectional design. It is a one-time study, not done multiple times over the years.
|
|
|
Naturalistic design
|
A naturalistic design is one where you observe people or actions in their environment in a natural, unobtrusive way. An experiment or a survey manipulates people or asks questions. Here is the great definition from about.com: "Naturalistic observation is a research method commonly used by psychologists and other social scientists. This technique involves observing subjects in their natural environment. This type of research is often utilized in situations where conducting lab research is unrealistic, cost prohibitive or would unduly affect the subject's behavior.
"Naturalistic observation differs from structured observation in that it involves looking at a behavior as it occurs in its natural setting with no attempts at intervention on the part of the researcher." |
|
|
Scatterplot
|
A scatterplot is a way of displaying numbers looking for relationships or correlations. Each item in a study becomes a data point on a graph. You are looking to see if as one thing gets bigger, what happens to the other dimension. Wikipedia's entry is useful. Frankly, the best way is to see the examples at the left what a scatterplot is. http://en.wikipedia.org/wiki/Scatter_plot
|
|
|
Linear regression
|
Linear regression is a big pair of words basically meaning the result of a scatterplot. Data are put in a chart and a scholar looks for a relationship between two sets of numbers. You are trying to see whether as one thing rises, does the other rise. In linear regression, you must add the "trend line" to the scatterplot and add the r.squared value. (Or r value.) What this shows is the trend line so you can predict future results. The r-squared value means "goodness" or fit. It is a correlational coefficient. The closer to 1 (or 1 and -1 for r values) the correlational coefficient is, the better fit your answer. The closer to zero, the less of a relationship there is. multiple regression. Multiple regression is a statistical concept. We don’t do multiple regression in this class, so this is just an introduction of concept. While linear regression compares only two variables, multiple regression tries to evaluate the interaction of more than two variables. Best wishes.
|
|
|
Likert scales
|
A Likert scale is a mathematical measurement of impression or of agreement. They are mostly used in surveys, and most people who have taken a survey have seen a Likert scale. A Likert scale is usually a scale as an odd number, a one to five scale, for example. The mathematics allows a research to do correlations and or means or other correlations. A typical Likert scale goes something like this: “To what extent do you agree or disagree with the following statement: “I like research.” 1 is strongly disagree. 2 is disagree. 3 is neither disagree nor agree. 4 is agree. 5 is strongly agree.
|
|
|
Spiral of Silence theory
|
Spiral of silence theory is a concept from communication science that suggests that when someone things they have a viewpoint that is in the minority, they will hide that viewpoint. It is especially important in a focus group to guard against the spiral of silence by asking open-ended questions or by asking people to make a commitment to an opinion (say by writing something down) before anyone speaks.
|
|
|
Informed consent
|
All participants in studies, experiments, surveys, and focus groups must give their informed consent. For surveys, informed consent may be as simple as an explanatory email. For depth interviews or focus groups, informed consent usually includes a signed form. Informed consent means: You tell them in general terms what the study is about. You tell them how their answers will be used. You tell them the degree to which they will remain anonymous. (You want to keep them anonymous). You tell them they don’t have to answer any question and that they may drop out at any time.
|
|
|
Institutional review
|
On any study involving humans, the institution must approve the parameters of the study through a process called institutional review. This process ensures that all scientific protocols are followed and helps a researcher see ethical problems that the researcher might not have noticed. At BYU-Idaho, institutional review goes through an academic vice president. There is an approval form on the university’s website.
|
|
|
Ratings
|
Nielsen Corporation developed ratings, a way to measure how many people watched particular television shows. Through a combination of set-top boxes and written diaries, Nielsen shows how many people are watching a show. A rating point means one percent of the households with a television set in a given area. In the United States, there are some 110 million households with TV. That means a one rating is 11 million homes. In Idaho Falls/Pocatello area, is you assume there are some 80,000 homes with television, then a 1 rating means 800 homes were watching that show. A rating is of all possible homes with TV, not necessarily those with TV on.
|
|
|
Households Using Television
|
Households Using Television, or HUT, is a ratings term showing the percentage of people in a given area using television. If each household had only one TV, then HUT for an area would be the sum of all of the ratings. However, since homes sometimes are watching more than one show at any given time, the sum of ratings may exceed the HUT.
|
|
|
Share
|
Share is the percentage of television sets in use watching a given show. So, in the early afternoon, a cooking show might get a 2 rating, but a 20 share. That would mean 2 percent of all TVs in the area were tuned into the show, but 20 percent of those households with TV on were watching that show.
|
|
|
Index numbers
|
Index numbers are a way of showing proportionally the degree to which a particular demographic group, let's say well-educated young women, will purchase a product or read a certain magazine. An index number of 100 is usually average. Here is how Jeff Hochstrasser describes it:
“Index numbers are most preferred for comparisons. They are the ratio of two percentages. “Generally, they are printed as whole numbers. An index number relates sales or product information to demographics. An index of 100 is usually average. Anything above is a percentage above and anything below is a percentage below average for usage. “Let us say that when we look at 25- to 34-year-olds, they represent 20 percent of the nation's population. In theory, they should represent 20 percent of the purchasing or consumption of a given product -- a 100 index number. Let's say they represent 20 percent of the population but 22 percent of the consumers of a product, say bottled water. We take 20 and divide by 22 and then multiply that produ |
|
|
SEO
|
SEO stands for search engine optimization. The concept is to learn how to research and develop websites that Google will find and put near the top of its rankings. Many resources are available. Google Analytics, Quintura are a pair of examples.
|
|
|
Literature review
|
The literature review is the part of the academic paper that sets up the research. It effectively shows what is known about the topic, or in other words, gives what is known about the meta-question. Effective literature reviews often detail holes in the literature, places where research can be done. Literature reviews often conclude with the research questions. They are usually written in third person. Some literature reviews sort of divide in several sections.
|
|
|
Abstract
|
The abstract is a 250-word summary of your paper. It appears right after the title page. An effective abstract makes it possible for a reader to review without reading the whole thing. Central findings and conclusions are included in the abstract.
|
