OOM Software Instructional Videos

"A learned, detached examination
of the well springs of modern
psychology."
Jude Dougherty
"Would you like to investigate
causality within psychology, as a
scientist might rather than a
statistician?"
Paul Barrett

OOM Instructional Videos...

Below you will find a variety of instructional videos for the Observation Oriented Modeling (OOM) software. A good place to start is with the Introductory Videos which explain the ideas behind the software and provide an overview of the program's structure and functionality. In separate tables below these two introductory videos you will find more specific videos organized into two categories: General Videos and Specific OOM Analysis Videos. You will also find a table of OOM Equivalent Analyses for Traditional Statistical Procedures such as t-tests, chi-squares, and ANOVAs. This latter table will help you to see how traditional statistical analyses can be replaced by OOM analyses.

Last Update: March 3rd, 2022; OOM Website



Introductory Videos

Watch these two videos to gain an understanding of the guiding principles behind OOM as well as the overall structure of the Windows program.

Video Title Description
OOM: What is it? How does it work? Why should I use it? As the title indicates, this video explains the general theory and philosophy guiding OOM. It also briefly explains how OOM differs from traditional statistics and shows a number of example analyses.  (~15 minutes in length)
 
Getting beyond NHST (the p-value) Why should psychologists and other social scientists abandon the common p-value (NHST)? Briefly, NHST is used to draw inferences to population parameters and requires random sampling to be successful; yet, most psychologists and social scientists do not even attempt to draw random samples. They are therefore clearly seeking a different inference, and in this video we argue it is the Inference to Best Explanation that is being sought. OOM provides the means for seeking such an inference.  (~10 minutes in length)
A first look at the OOM software Provides a brief preview of the software, showing how analyses are focused on patterns and persons (observations) rather than upon means, standard deviations, and variances.

 


General Videos

These videos offer general information about OOM's structure and functionality. For example, here you'll find videos about how the program installs to your computer, how to run OOM on a Mac, and how to navigate the Data Edit and Text and Graphics Output windows.

Video Title Description
Installation & Overview Explains how OOM installs onto your PC. The User's Manual installs with the program in Word and Adobe Acrobat format, and a large number of OOM example data files install automatically as well.  This video also provides a general overview of the OOM software's structure and functionality.  (~5 minutes)
Running OOM on a Mac Demonstrates how to use Winebottler, a free program, to run OOM on a Macintosh computer. There is no separate Mac version of OOM, so the program must be run through a PC emulator such as Winebottler. (~ 9 minutes)
Data Edit Window Toolbar Explains the functions of the various tool buttons on the toolbar of the Data Edit window.  (~ 5 minutes)
Data Edit Window Sub-Toolbar Explains the functions of the various tool buttons on the sub-toolbar of the Data Edit window. (~ 5 minutes)
Text Output Window Describes the different features in the Text Output window, including how fonts can be changes and how the number of decimals in numeric output printed can be changed. (~ 6 minutes) 
Graphics Output Window Describes how graphs are created, viewed, edited, and saved in OOM. Also demonstrates how graphs can be exported to MS Powerpoint for editing. (~ 6 minutes)
Data Entry #1
Data Entry #2
Data Entry #3
These videos demonstrate how data are entered and labeled in the OOM software. Labeling units of observation is very important in OOM as this is tantamount to determining the Deep Structure of the data. Most analyses are based on the Deep Structures of the data.  These videos also include descriptions of the Define Orderings window and the Auto Generate options for defining units of observation. (~4 to 8 minutes in length)
Deep Structure Explains the concept of Deep Structure in OOM and gives several examples. (~ 5 minutes)
Importing from Excel
Importing from SPSS
Demonstrates how data can be imported from an Excel spreadsheet or from an SPSS file. Each video is ~6 minutes in length.
Generate Data Shows how data from proportions or frequencies (e.g., as reported in a contingency table) can be created quickly using the Generate Observations option.
Multigram Options
Editing Multigrams
The first video describes the various display features for multigrams in the OOM software. The second video shows how multigrams can be exported as Windows Metafiles and then edited, in detail, in other programs such as Powerpoint and Word.

 


Specific OOM Analysis Videos

These videos demonstrate the primary analysis features of the OOM software such as the Build/Test Model, Ordinal Analysis and Efficient Cause Analysis options.

OOM Analysis Video Analysis Description Related Materials
Build/Test Model

 

The Build/Test Model option allows you to explore the relationships between two or more orderings (variables). For example, how are aspirin ingestion and the common cold related, or how is relationship investment and satisfaction related to relationship commitment? The relationships are not assumed to be linear, and the analysis uses a Bayesian-like classifier to determine how many individuals fit the discovered pattern (if any) within the data. The Build/Test option thus offers post hoc methods for identifying patterns; in other words, you are not required to specify an expected pattern prior to analysis. The primary visual tool is a multigram, a number of examples which can be viewed here.  
Ordinal Analysis: Concatenated Orderings This analysis allows you to examine ordinal differences across two or more orderings. This may be useful, for example, in analyzing changes in quantities over time, such as with a pre- and post-test study design. Normally, you will specify the predicted ordinal pattern before running the analysis. Here is an example defined pattern in which quantities are expected to decrease from pre-test to post-test:

     Pre-Test
     |  Post-Test
     |  |
     +  O Highest
     O  + Lowest    (+ indicates expected ordinal pattern)

Ordinal Analysis: Crossed Orderings This analysis allows you to examine ordinal differences across two or more groups of individuals. For example, suppose a memory researcher assigns participants to two experimental (A and B) and one control group. The outcome variable is number of words recalled from a memory task, and the researcher expects group A to recall more words than group B, and group B to recall more words than the control group. The expected ordinal pattern across the three groups is therefore: A > B > Control. Here is pattern in visual form:
 
     A  B  Control
     +  O  O   Highest
     O
     O   +   Lowest    (+ indicates expected ordinal pattern)
 
Pattern Analysis: Concatenated Orderings This analysis allows you to test a specific pattern of observations across several orderings. For example, you might consider binge drinkers to self-report relatively high levels of extraversion and agreeableness and relatively low levels of conscientiousness. If the personality scale is divided into four quarters, then the expected pattern may appear as follows:
 
         E  A  C
     Q4  +  +  O  
     Q3  +  O
     Q2  O  O  +
     Q1  O   +      (+ indicates expected locations of observations)
 
Pattern Analysis: Crossed Orderings This analysis allows you to test a specific pattern of observations within a two-dimensional matrix. This may be useful, for example, in comparing two groups on a rating scale. You must specify the predicted pattern before running the analysis.  Here is an example defined pattern:

  Scale Rating
     1 2 3 4 5 6
     + + + O O O Group 1
     O O O + + + Group 2   (+ indicates expected rating)
Grice, 2015 publication demonstrating this analysis.
Efficient Cause: Blocked Orderings Efficient causes are those ordered in time, as with the classic S --> R concept in psychology. The Blocked Orderings efficient cause analysis can be used to analyze data that are blocked across time, like those found in an ABAB research design. For example, a behavioral clinical psychologist might examine the causal efficacy of a treatment program by assessing the number of disruptive behaviors for each child in a classroom. She assesses five time points within each of four blocks of trials in an ABAB design: 1) no treatment, 2) treatment, 3) no treatment, and 4) treatment. She might expect the following ordinal pattern of disruptive behaviors across the four trials: 

     A
     |  B
     |  |  A
     |  |  |  B
     |  |  |  |
     +  O  O  O Highest # disruptive behaviors
     O  O  +  O
     O  +  O  O
     O  O  O  + Lowest # disruptive behaviors    (+ indicates expected ordinal pattern)

 Each child's data can also be examined and assessed with this analysis.

Parker and colleagues review methods similar to those in OOM:
     

 


OOM Equivalent Analyses for Traditional Statistical Procedures

     Select instructional videos are organized here according to their comparisons with traditional statistical procedures such as t-tests, chi-square, and ANOVA.
 

Traditional Statistical Procedure Analysis Goal OOM Analysis Video
Independent samples t-test Compare two independent groups regarding an observed quantity; e.g., compare men and women on their observed levels of introversion, or compare experimental and control groups with regard to their number of eye-blinks while watching a video clip.
Dependent samples t-test Compare two dependent sets of observations on an observed quantity. For example, who (on average) is more satisfied in marriage? Pairs of husbands' and wives' scores on the Dyadic Adjustment Scale could be compared to answer this question. Each pair of scores are clearly dependent. Quantitative data from a  pre-test/post-test research design could also be analyzed. 
Between-Subjects ANOVA Compare quantities for two or more independent groups; for example, examine difference in IQ scores between left-handed, right-handed, and ambidextrous individuals. 
Repeated Measures (or Within-Subjects) ANOVA Compare quantities across two or more occasions or time points; for example, examine changes in heart rate across four different measurement occasions in an experiment. Quantities with identical ranges from different questionnaires or scales can also be compared; for example, Big Five trait scores can be compared with regard to their relative magnitudes within individuals.
Chi-Square Test of Association (or Independence) Are two categorical variables associated (correlated)? The Chi-Square Test of Association addresses this question. For example, is aspirin consumption (aspirin vs. placebo) associated with a reduced incidence of a second heart attack (attack vs. no attack) among elderly men?  
Chi-Square Goodness-of-Fit Test This analysis is used when the researcher has one categorical variable and wishes to examine how observations are distributed across that variable. For example, a dice manufacturer may wish to test a six-sided di for fairness. If the di were rolled 1002 times, the expected frequencies of 1's, 2's, 3's, 4's, 5's, and 6's would each be 167 (1/6 * 1002). Do the actual rolls of the di match expectation (at least closely)? The chi-square goodness-of-fit test can be used to evaluate this question. 
Single-Subject or small n Analyses (e.g., for ABAB designs) Many study designs involve administering and withdrawing an intervention or several interventions over time. Each administration or withdrawal of an intervention is referred to as a block, and the researcher will be interested in examining the impact of these changes on an observed outcome. For example, a behavioral clinician might study the effectiveness of an intervention program in reducing disruptive behaviors among children.

More to come...

   

 


Return to the OOM website