Psych 2820 Midterm 1 Review

21 Pages
Unlock Document

Management and Organizational Studies
Management and Organizational Studies 1021A/B

Midterm 1 Review 10/22/2013 The Scientific Method and Psychology - psychology is a science which has its goals provide explanations of behavior and experience - research that adheres to establish methodological principles and procedures, and the results of that research are analyzed using statistical procedures - these methodological procedures and statistical techniques exemplify what is known as the scientific method - qualitative research generally deviates from the scientific method -GordonAllport stated that the aim of psychology should be, “enhancing-above the levels of achieved by common sense-our powers of predicting, understanding and controlling human action” - In its infancy, Psychology used an investigation method called introspection. Introspection involved having individuals report on their subjective impressions of various forms of stimulation. (Ex. They might report if how they were able to distinguish one color from another color) - Psychology struggled to gain acceptance as a real science while introspection was its main investigation tool - Another problem that stopped them from being accepted as a real science was the fact that it claimed to study consciousness, something that could not directly be observed, and thus was outside the realm investigation by the methods used in the physical sciences - If psychology was to be a real science it needed to follow the guidelines that were apart of the Positivist Movement (associated withAuguste Comte). The positivists suggested that if a discipline was to be viewed as scientific then its data must be derived from publically verifiable, objective observations, its results must be replicable, its explanations must not incorporate metaphysical statements about the nature of events, and one of its goals should be the attainment of practical knowledge. - Some of the important characteristics of scientific explanations derived from the positivist viewpoint are:  Lawfulness: every event (behavior) can be understood as a predictable sequence of natural causes to effect Determination: behavior is solely influenced by the natural causes and not by unknowable forces, such as “free will”  Using objective, systematic, and publically verifiable observation and recording of events as data. Scientific explanations are based on facts derived through empirical enquiry. Facts are events that can be directly, objectively, and repeatedly, observed thereby resulting in publicly verifiable data. Empirical inquiry is the essential characteristic of the scientific method  Naïve Empiricism or Strict Positivism: if something is not sensed it does not exist. Ex. Gravity – this is another important issue b/c many of the explanatory concepts used in psych are not directly observable. - Asolution to this problem was found in the philosophy of Logical Positivism, which is associated with the Vienna Circle (circa early 1900s) - a group of physicist, logicians, and mathematicians who wanted to formulate the principles for gathering knowledge. Logical Positivism or Sophisticated Empiricism is based on inference, which is an intellectual process in which conclusions are derived from observable evidence - Constructs: unobservable phenomena Operationalize or giving an operational definition: unobservable theoretical constructs could logically be tied to observable events, and thus the unobservable construct could be manipulated and measured by reference to the observable event S.S. Stevens: devised the scale of measurements, nominal, ordinal, interval, and ratio - Constructs are analogical tools for organizing and trying to understand the operation of observable phenomenon Reification of a Construct: elevating a construct to the level of reality – they are so frequently used that people start to treat the construct as a real entity  Awillingness to change previous beliefs on the basis of new information – science is a progression Falsifiability: the idea that an explanation of behavior must be testable to be useful - Karl Popper argues that falsifiability is an essential part of the scientific method - False predictions lead to refinement of reformulation of the theory  Scientific explanations should conform to Occam’s Razor – in explaining behavior the simplest theories be preferred - C. Lloyd Morgan – we should avoid making any more assumptions than absolutely necessary to explain animal behavior  Empirical Principles or Law: assertions accepted as truths on the basis of empirical inquiry. Laws do not have to state cause – effect relationships, they mat simply state association  Inductive Reasoning: specific to general Deductive Reasoning: general to specific Hypothetico-Deductive method: research is guided by this: 1. Hypothesis 2. Designing research to gather data 3. Determining how well the data conforms to the hypothesis 4. Re-evaluating the theory Theory: an analytic structure, comprised of principles and laws, that seeks to explain a set of observations Hypothesis: specific statements about how some variables should affect or be related to other variables Research may be viewed as involving several sequential steps: ­ generating idea ­ literature review – refinement stage o deductive reasoning is paramount in this stage ­ how the research should be conducted – methods and design ­ data collection ­ analyze data ­ interpret outcomes o inductive reasoning is of importance in this phase Different Types of Relationships: Descriptive, Correlational, and Causal It is often important to have a description of some behavior ­ If the behavior is a novel type of behavior, it is important to know things like how often does it occur, who performs the behavior, what is the average amount of behavior, and does it exhibit a wide range of values or only occur within a narrow range of values ­ Studies are conducted to examine these variables in a sample (subset of population) of participants, and those sample statistics are used to make statements about the population parameters. One way this might be done would be to firm a confidence interval. ­ Even if a behavior has been occurring for a long time, it is often important to know if that behavior has changed ­ Descriptive studies are useful because they allow researchers and other consumers of the research to use that information to make decisions. ­ Descriptive studies, in which some aspect of a variable are measured, do not allow the researcher to conclude anything about what other variables may be related to or causing any change in the measured variable. ­ Descriptive studies, in which a variable is simply measured, even if the measurements occur in several different subgroups are, as the name implies, purely descriptive. Once an interesting and important behavior has been identified and described psychologists design studies that will lead to a better understanding of the behavior. Two key components in the understanding of the behavior are identifying the variables that are related to the behavior and identifying the causes. These two key components are the basis of two major types of relationships that research may be designed to identify: correlational relationships and causal relationships. ­ In a correlational relationship, the values of two (or more) variables covary, which means that changes in the values of one variable are accompanied by reliable and consistent changes in values of another variable. However, the fact that two variables tend to covary does not mean that one causes the other. ­ Post hoc ergo proptor hoc fallacy: one of the most common errors made by people not versed in research design and statistics is to misidentify a correlational relationship as a causal one. o Personal biases can often lead to eagerness to see causation in correlation - Much research in psychology is used to determine correlation, because if it is known that two variables covary, then it is possible to predict the value of one of those variables if the values of the other variable are known. - Using the existence of a correlational relationship as the basis of prediction involves what is called regression analysis. In regression analysis , one of the variables is used to make predictions about the values of the other variable - The variable that is being predicted is called the criterion variable and the variable from which the prediction is made is called the predictor variable. There are three potential problems that make it difficult to determine causality from the existence of a correlation: 1. Third variable problem: two variables may covary because both of them are related to a third variable. This problem leads to what is called a spurious correlations, two variables are related only because they are both related to another variable 2. The existence of a correlation does not provide any information about the potential direction of the relationship. This is called the directionality problem. 3. Selection Bias can also lead to spurious correlations. In this context, selection bias refers to the fact that individuals with certain characteristics may be more likely to choose different activities, environments, etc. and this may then lead to a spurious relationship between those individuals and some other variable (this is like the third variable problem but is given a separate treatment here). - Acausal relationship exists if changes in one variable produce changes in a second variable. Covariation rule: correlation or covariation is a necessary condition for causation, but is not a sufficient condition. -The identification of causal relationships is the purview of true experiments - There are two conditions that must be present in a research study in order for the study to have the ability to determine if a causal relationship exists between two variables – that is, in order for the study to qualify as a true experiment: 1. Use of a true independent variable(s) - studies that want to identify causal relationships involve the use of groups of participants that differ on the basis of receiving different treatments. - Experimental Group: one group of participants that might receive some treatment - Control Group: the other group of participants that would receive no treatment - If the researcher can manipulate the different types of treatment so as to give any of the treatments to any of the participants then the grouping variable is called an independent variable. - The use of independent variables to form the groups in a study is one of the keystones of a true experiment – a study that has the potential for identifying causal relationships. - Static group variables, participant characteristics, or quasi-independent variables are variables which categorize research participants and which can form the basis or group formation, but which cannot be manipulated. These variables can be used to place participants into naturally-occurring groups. 2. Random assignment of participants to different levels of the independent variable - Another requirement for a study to be a true experiment is that the researcher must be able to randomly assign participants to the different levels of the independent variable. The purpose of an experiment is to determine if the manipulation of the independent variable causes some change in the behavior under study. The behavior under study is called the dependent variable. Random assignment helps to eliminate the influence of extraneous variables. Internal and External Validity Internal Validity In order to unambiguously ascribe changes in the dependent or criterion variable to either the independent variable (true experiment) or predictor variable (correlational study), it must be possible to rule out plausible alternative explanations of the relationship. This is what is referred to as internal validity. In an experiment, this means showing that variation in the independent variable caused the observed variation in the dependent variable. Internal validity is directly affected by how successful the researcher has been in eliminating the effects of confounding variables that allow for alternative explanations. Aconfounding variable is an extraneous variable (any variable which is not what the researcher envisions as the causal or predictor variable) that systematically covaries with the variable under study and thus could also account for the observed outcome. Minimal experimental design: using at least two groups formed on the basis of different levels of an independent variable and randomly assigning participants to the groups. Threats to Internal Validity: History - This refers to events, external to the participant, not related to the variables under study, that take place between measurements. - History effects are of special concern in studies with multiple observations separated by substantial time since events occurring between observations may influence the results - The longer the time in between measurements the greater the possible effect of history Maturation - Events that change over time that are internal to the participant - Maturation as a threat to internal validity is particularly important in studies of development Biased ParticipantAssignment - Selection bias occurs if participants in one group differ initially from those in another group. - Selection bias may occur because the researcher does not or cannot randomly assign participants to groups Testing - With multiple testing, subsequent test results may be influenced by previous test results Instrumentation - In studies involving repeated measurements if changes are made in measurement instruments or observers, there is no guarantee that subsequent measurements are comparable to the initial measurements - The method of collecting the dependent measures or criterion variables should not be changed during the course of a study Statistical Regression to the Mean - When participants have been selected on the basis of extreme scores, regression to the mean will occur - Regression to the mean was first recognized by Sir Francis Galton in his study of physical and mental characteristics of humans - Reversion to the mean: extreme scores at time one tend to become less extreme at time two - Adistinction between the true value of some variable and the currently observed value of the variable - True values of a variable are relatively stable, while a currently observed value of the same variable will be influenced by a variety of chance and random variables (introducing what is called error variance) - Further understanding of statistical regression is possible by using Z scores AZ score is a measurement of how far a value is away from the mean. The distance is expressed as the number of standard deviations - In understanding regression to the mean it is helpful to know that the relationship between a measure taken a second time is related to the measure taken a first time by the formula: Zy = rxy Zx. Zy = the measure taken the second time Zx = the measure taken the first time The second measurement (Y) must always be closer to the mean - Astrong correlation indicates less error variance in your measurement, the measurement is a more accurate reflection of the true value - Regression to the mean is of particular concern in pretest-posttest designs that do not include a control group ExperimentalAttrition - When participants do not complete the study you must be concerned with whether the attrition is random or selective - If ‘drop out’is random, then internal validity would not be compromised by participant attrition (although the power of the study would be detrimentally affected) - If attrition is selective, i.e. only participants with certain characteristics drop out, then internal validity is compromised External Validity - External validity relates to the generalizability of results - To the extent that the results of a study can be said to generalize to the people (sometimes called population external validity) and conditions ( sometimes called situation external validity or external validity with regard to setting) that are the object of the study, the study is said to have good external validity Factors that threaten or limit external validity: Biased Participant Selection - If the sample is not representative of the population of interest then the results obviously would not generalize Participant attrition - Even if you start with an unbiased and representative sample if there is nonrandom participant attrition you will be left with a biased sample that would not generalize well to the originally sampled population Measurement Reactivity - Any procedure that could possibly make the participant react differently than he/she normally would The external validity of studies can be increased if the study is designed to be high in mundane realism or experimental realism. - Mundane realism refers to making studies as similar as possible to the real world.Achieving mundane realism is sometimes impractical, often for financial reasons, so it has been suggested that studies be high in experimental realism. The term ecological validity if often used to describe studies that are conducted in more naturalistic settings. - Experimental realism refers to the degree to which participants experience psychological states in which the researcher is interested in and thus behave in a natural manner. In many cases, the production of experimental realism requires some level of deception Often internal and external validity are at odds. In most psychological research, external validity is often treated as secondary to internal validity. There are, however, some types of qualitative research where external validity is considered more relevant than internal validity. Pre-experimental and Quasi Experimental Designs The true experimental design is used as the mode against which other designs are compared.A true experiment has the following essential characteristics: - a least two groups - manipulation of a independent variable - random assignment of participants to groups - attempts to control extraneous variables and to increase internal validity Pre-experimental and Quasi-Experimental: refer to mainly field studies that are well-designed and as much as possible exert good control over extraneous variables, but which lack some of the essential features of a true experimental design True Experiment VS Pre-Experimental or Quasi-Experimental - manipulation of independent variable under control of researcher VS presences of some treatment or intervention but often not under researcher’s control - random assignment to groups VS naturally occurring groups - good control over extraneous variables VS less control over extraneous variables, but attempts are made to control - ideal for researching causal conclusions VS cannot reach causal conclusions, can only study correlational relationships - One common aspect of quasi-experimental designs is that participants are not randomly assigned to treatments, rather they enter the study based on being part of some naturally- occurring group.
More Less

Related notes for Management and Organizational Studies 1021A/B

Log In


Join OneClass

Access over 10 million pages of study
documents for 1.3 million courses.

Sign up

Join to view


By registering, I agree to the Terms and Privacy Policies
Already have an account?
Just a few more details

So we can recommend you notes for your school.

Reset Password

Please enter below the email address you registered with and we will send you a link to reset your password.

Add your courses

Get notes from the top students in your class.