click below
click below
Normal Size Small Size show me how
Chapter 7 RM
| Question | Answer |
|---|---|
| Experimental | manipulation when experimenter treats group differntly according to IV levels, Random assignment to Ps to conditions |
| Correlation | No mainuplated variable, no independent variable |
| Simple Experiement | IV is manipulated, DV is measured |
| Basic Experiment | 2 levels of IV, experimental group, control comparison group |
| Control over extraneous variables | randomizing, matching |
| * | Casual relationship between IV and DV can be established |
| Methods to contol confunding variables | Design to experiemtn, randomization, matching, control of partcipants effects/ experimenter effects, manipulation check |
| Randomization | most important |
| Randomization selcection | Representativenesss, eliminates systematic bias, increase external validity |
| Random assignment to conditions | equal distribution of partcipant characteristics to conditions, increases internal validity, more important than random selection |
| Value of assignment- manipulating the treatment | random assignment makes the IV the obly systematic difference between groups, no random assignment= no experiment |
| Control procedures | increases sensitivity of an experiment, eliminates that variable are being controlled as a cofund |
| Holding variables construct | creates a more homogenous sample |
| Matching by building extraneous variables into the research desing | to compare two populations |
| Building extraneius variables into the design | excellent for acheiving control and increasing sensitivity, matching |
| Matching by Yoked Control | controls the possible influence of the time relationsjip between an event and behavior |
| Control of parcipant effect | demand characteristics, double blind placebo model, deception, manipulation check |
| Control of experimenter effects | practice beig consistent, run all conditions at the same time, automation, recorded instructions control recording errors- procedural reliability blind to condition control of exoeriments attributes error- blind counter balance |
| Remember that | Simple experiment is a simple way to establish internal validity, independent random assignment = cornerstone, control partcipant diffrence through matching, hypthesis testing to determine if the differences are relaiable. Type 1 vs type 2 error- t-test |
| Reporting the T-Test results | Include statistical test, degress of freedom, probability of significance level- if significance level is below .05 you can reject the null hypthesis. if above .05 you retain the null hypthesis |
| Results- inferential statistics | means, standard deviation and standard erro for each measure for each group report the inferential statistic results, explain results, give statisical evidnece, develop and claify the point use impartial scientific reporting |
| Assumptions of the T-Test | observations are indenpendent, internal or ratio, underlying distributions are normally distributed both groups distributions have the same variance |
| Analyzing the results of the simple experiment | decide on significance level before doing experiment- is the difference not due to change, if not statistically signicant- conclude nothing, if statisically significant- effect probably reliable but not neccessarily large/important |
| Power- reducing type II error | reduce the effect of random error, standarize procedures and use relaible measures, code data carefully |
| Statisfically significant results- declaring that the treatment has an effect | not necesairlily large control group rule out- placebo effect, experiment bias and other confunding variables |
| Statisical question | is the observed mean difference greather than what is expected based on the assumed diffrence of zero |
| Null Hypotheis | the treatment doesnt have an affect |
| Experimental hypthesis | treatment has an effect IV-> DV |
| Simple Experiment | Overview- errors in determing weather results are statiscally significant statistic and the design of the simple experiment, nonstatistical considerations in the design of the simple experiment, analyzing data from the simple experiement |