Research Design, Statistics, Tests, and Measurements Flashcards
Nonequivalent Group Design
In a nonequivalent group design, the control group is not necessarily similar to the experimental group since the researcher doesn’t use random assignment. This is common in educational research because you can’t randomly assign subjects to different classes. For example, if a researcher wants to see if a new method for teaching reading is better than the usual method, the researcher might assign the new method to one class and the usual method to another class, and measure each subject’s increase in reading skill from the beginning to the end of the study.
External Validity
The degree to which the results of an experiment may be generalized.
Internal Validity
The certainty with which results of an experiment can be attributed to the manipulation of the independent variable rather than to some other, confounding variable.
Hawthorne Effect
The alteration of behavior by the subjects of a study due to their awareness of being observed. One way to control for the Hawthorne effect is to use a control group design and observe both the control group and the experimental group.
Single-Blind vs. Double-Blind Experiments
In a double-blind experiment, neither the researcher who interacts with the subjects nor the subjects themselves know which groups received the independent variable or which level of the independent variable. If the subjects do not know whether they are in the treatment or control group, but the researchers know, it is called a single-blind experiment.
Placebo Effect
A beneficial effect produced by a drug or treatment, which cannot be attributed to the properties of the treatment itself, and must therefore be due to the patient’s belief in the treatment. The placebo effect is a special kind of demand characteristic. One possible remedy for the placebo effect is to have control groups, so that the effectiveness of the drug over-and-above the placebo effect can be determined.
The 2 basic types of statistics
Descriptive statistics and inferential statistics. Descriptive statistics is concerned with organizing, describing, quantifying, and summarizing a collection of actual observations. With inferential statistics, researchers generalize beyond actual observations. That is, inferential statistics is concerned with making an inference from the sample involved in the research to the population of interest, and providing an estimate of population characteristics.
Frequency Distribution
A frequency distribution is a graphic representation of how often each value occurs.
3 measures of central tendency
The mode, the median, and the mean all provide estimates of the “average” score.
The Mode
The mode is the value of the most frequent observation in a set of scores. If two values are tied for being the most frequently occuring observation, the data has 2 modes, or is bimodal. A distribution can also have 3 modes, 4 modes, etc.––or no mode, if every value in the distribution occurs with equal frequency. This makes the mode different from the other two measures of central tendency, as there can only be one median and one mean.
The Median
The median is the middle value when observations are ordered from least to greatest, or from greatest to least. If there are an even number of data points, then the median is the arithmetic mean of the two middle-most numbers.
The Mean
The mean of a data set is the sum of scores divided by the number of scores. The mean is the measure of central tendency most sensitive to outliers, i.e. extreme scores. If you have outliers in your data set and you are interested in a representative score, it usually makes sense to use the median, and not the mean, as your measure of central tendency.
Percentile
The percentile tells us the percentage of scores that fall at or below a particular score.
z-score
A z-score expresses how many standard deviations above or below the mean a particular score is. To determine z-scores, you subtract the mean of the distribution from your score, and divide the difference by the standard deviation. Negative z-scores fall below the mean, and positive z-scores fall above the mean.
Mean, median, and mode in normal distributions versus skewed distributions
Because the normal distribution is symmetrical and has its greatest frequency in the middle, the mean, median, and mode of a normal distribution are identical. In skewed distributions, where the distribution of scores is not symmetrical, the mean, median, and mode are not identical.
T-scores
Don’t confuse T-scores with t-statistics. Z-scores can be converted to T-scores. The T-score distribution has a mean of 50 and a standard deviation of 10. So, for example, a T-score of 60 is one standard deviation above the mean. Because of their nice round numbers, T-scores are often used in test score interpretation.
Correlation Coefficients
Correlation coefficients are a type of descriptive statistic that measure the extent to which two variables are linearly related. Correlation coefficients range from -1.00 to +1.00. A positive correlation means that as the value of one variable increases, the value of the second variable tends to increase as well. A negative correlation means… The absolute value of a correlation coefficient tells us how strong the relationship is. If two variables have a correlation of zero, knowing the value of the first variable does not help you predict the value of the second variable. The graphical representation of correlational data is called a scatterplot.
Best-Fitting Straight Line
Used to highlight correlation on a scatterplot.
Factor Analysis
See the physical flashcard.
Inferential Statistics
Inferential statistics is concerned with making inferences, or generalizations, from samples to populations, while taking into account the possibility for error.
When should the criterion of significance be chosen?
Prior to collecting data.
Significance Testing Process
See physical flashcard.
Errors in Significance Testing
There are two possible errors. A Type 1 error is when you reject the null hypothesis by mistake. The likelihood of making a Type 1 error is called alpha, and is the same as the criterion of significance. A Type 2 error is when you accept the null hypothesis even though the null hypothesis is false. In other words, you obtain a statistically insignificant result and conclude, wrongly, that the null hypothesis is true. The probability of making a Type II error is called beta, and depends largely on sample size and variance. (Note that the probability of making a Type 1 error is called alpha and the probability of making a Type 2 error is called beta.)
Note regarding significance testing
The purpose of significance testing is to make an inference about a population on the basis of sample data. Statistical significance does not tell us anything about whether the research is poorly designed, or whether the results are trivial or meaningless. The larger the size of the sample, the smaller the difference between the groups has to be in order to be significant. Therefore, if you use really large sample sizes and you get a statistically significant result, the difference between the groups on the DV measure might be so small as to make the results trivial.
F ratio
See physical flashcard.
Factorial Design
Each level of a given independent variable occurs with each level of the other independent variables.
Meta-Analysis
Meta-analysis is a statistical procedure that can be used to make conclusions on the basis of data from different studies. If researcher A publishes a study on therapeutic outcomes, and researcher B publishes a similar study using different methods, we can use meta-analysis to combine the results of these studies and come up with a more general conclusion.
Face Validity
Face validity refers to the degree to which a procedure appear__s to measure what it is supposed to measure. If you are interested in measuring knowledge of 20th-century American history, but you give subjects a test on 20th-century European history, the test will lack face validity.
Scales of Measurement
See physical flashcard.
The 2 types of ability tests
There are two types of ability tests: aptitude tests and achievement tests. Aptitude tests are used to predict what one can accomplish through training. In other words, they are used to predict future performance. Intelligence tests are aptitude tests. Achievement tests, on the other hand, attempt to assess what one knows or can do now.
Personality Tests and Inventories
Formally, in taking a test, the subjects are instructed to do their best; in completing an inventory, they are instructed to represent their typical reactions. A personality inventory is a self-rating device usually consisting of somewhere between 100 and 500 statements. The subject is asked to determine if the given statements apply to him or her. Although these structured tools are quite reliable, the veracity of responses is not guaranteed. For example, if an item says, “I occasionally steal,” most people will tend to answer “no” regardless of whether or not they occasionally steal. The perceived social acceptability of a response is just one factor that can affect the accuracy of inventories that involve self-reporting.
How the Minnesota Multiphasic Personality Inventory (MMPI) was first revised
In 1989, a revision of the MMPI, the MMPI-2, added content scales. These scales were formed using items derived from theoretical concerns rather than from an empirical criterion-keying approach. For example, to form the low self-esteem content scale, the authors selected items that ought to be related to low self-esteem. Hence, the original clinical scales have been supplemented with content scales that were developed using a more theoretical approach.
Minnesota Multiphasic Personality Inventory (MMPI)
The MMPI is one of the major personality inventories. It consists of 550 statements to which subjects respond “true,” “false,” or “cannot say.” The MMPI yields scores on ten clinical scales, measuring things such as depression, schizophrenia, and masculinity/femininity. It has scales that can indicate whether the person is careless, faking answers, misrepresenting him- or herself, or distorting responses, and whether the distortion is being done intentionally or unintentionally. The purpose of the MMPI is to aid in the assessment of various clinical disorders. All scores on the MMPI are expressed as standard scores with a mean and standard deviation derived from standardization samples.
Standardization Sample
A standardization sample is a population of individuals who have previously well-documented intelligence and/or achievement levels, which is used to “standardize” new or revised test instruments to assure that they are reliably measuring what they are intended to measure.