Psychology 2011 1997 Lab Week 6 Summary

These pages are demonstration pages of mixed mode deli very for an on-campus course. The curriculum and course structure has changed substantially for 1998, but these pages have been retained here as an example of a particular use of the Web in teaching.

Apologies for no weekly summary last week on Week 5 material. Please refer to the Introduction Section site of the Long Report Writing Guide on the Web.

WUG Data and Results Sections

The summary below is divided into two sections. The first section is a copy of the Results proforma you were required to complete last week with suggested answers to each question. The second section outlines some points on the presentation of results in long reports this year. These points will be re-emphasised again in labs this week.

WUG RESULTS PROFORMA (example answers provided)

1) Are there differences in level of performance for the different variants of the regular plural and the different variants of the past tense? (Compare the number correct in each type)

For this question you should be looking at the mean correct (and SEs) for simple plural vs complex plural, and simple past-tense vs complex past-tense from the entire monolingual sample. Remember, the means and SDs presented in the first two tables within Experiment one are from a subset of the entire monolingual sample and won't be telling the whole story.

For plural morphology, performance on the simple allomorph (Mean = 4.75, SE = 0.12) is better than performance on the complex allomorph (Mean= 2.99, SE = 0.29). For past-tense morphology, performance on the simple allomorph (Mean =1.29, SE= 0.09) is better than performance on the complex allomorph (Mean = 0.92, SE=0.08).

2) Is performance equivalent for the plural and past-tense cases? (You will need to compare percentages correct, since there are different numbers of cases)

You can work out a combined percentage for plural and compare it to the combined percentage for past case, but it would be more accurate to compare simple plural with simple past-tense, and complex plural with complex past-tense. Performance is superior for the plural cases.

3) Where the child does not produce the same form as an adult would, is there a predominance of a particular error type? Does the pattern differ for plural versus past-tense?

Try to be specific here, and describe the pattern of errors. Do not just look to see which mean is higher and state that is the predominant error type.

For plural morphology, when the child does not produce the same form as an adult would, they predominantly produce the zero allomorph. For past-tense the common error-types are zero and "other" (means between zero and ‘other’ are too close to say either one is predominant).

4) Describe and interpret the analysis of developmental trend. (You will need to refer to the appropriate statistics in your sentence (s))

This question is just looking at the effect of age on allomorph production.

From Table 1 it appears that older children are performing better than younger children on plural morphology. This is supported by a significant main effect of age, F(1,58)=8.07, p, .05. A similar pattern is observed in Table 2, with older children performing significantly better than younger children on past-tense morphology, F(1,58)=8.92, p< .01.

Leave any further interpretations of the data patterns to the discussion.

5) Describe and interpret the analysis of language history. (You will need to refer to the appropriate statistics in your sentence(s))

This question is just looking at the effects of language history on allomorph production.

From Table 3 it appears that monolingual children are performing better than bilingual children on plural morphology. This is supported by a significant main effect of language history, F(1,23)=6.58, p< .05. From table 4 it appears monolingual children are also performing better than bilingual children on past-tense morphology, however, this apparent difference between the two groups was not significant, F(1,58)=3.60, p> .05.

Hints and Tips for writing Results Sections

Intially you must describe how you treated the data and what form the data to be presented will take. Two key aspects of describing data treatment are:

what (if any) transformations have taken place?
are any of the data collected omitted and if so, why?

A general statement of the important aspects of the data is required. This statement should refer to a figure or table (present tense when referring to figure), making it clear what exactly has been graphed or tabled. The figure and table captions are not part of the text, and so the description of the figure or table should not assume that the caption has been read. Any differences or key elements of the results should be highlighted, and should be defended by giving summary statistics (or making reference to the table or graph) and stating results of any inferential statistical tests performed (whether significant or not). For example:
Examination of Figure 1 shows that both younger and older subjects performed significantly better on the simple plural items compared to the complex plural items, F (1,46)=45.44, p< .001.

Important Points

Defend statements with reference to:

Summary statistics
- Table of means with standard deviations
  OR
- Figure with standard error bars
Inferential statistics
Indicate if significant or not (substantiating description of data)

Reporting Statistics

Italicise all statistical symbols in standard alphabetic letters (e.g., t, p, F, M )
For any inferential statistic give:
- name of statistic t
- associated degrees of freedom t(26)
- equals sign t(26) =
- obtained value t(26) = 2.66
- probability level t(26) = 2.66, p<.05
Separate inferential statistical information from your general statements with a comma or brackets.
Report percentages as whole numbers; other numbers give two decimal places
Do not give lead in zeros (e.g., 0.50) for probabilities or correlation coefficients

Example description:
Table 1 shows that the increase in mean confidence ratings was significantly greater in the experimental group (M = 5.01, SD = 3.52) than in the control condition (M = 3.84, SD = 3.02), t(26) = 2.66, p<.05.

Note use of present tense to talk about table (which reader can see) and past tense to talk about the results.

Common problems in Results sections

Starting with a table or figure
Table/figure captions lacking descriptive information or are misleading
Remember there is a huge difference bewteen Insignificant and Non Significant results
To much emphasis on statistics and not enough on describing the data (don't just jump into the reporting of inferential statistics)
Reporting of statistics inconsistent with description of trends in data, or presentation of data in figures (watch for inappropriate axes scales that exaggerate or reduce actual differences described in text).

For more information please use the Results Section from the Long Report Writing Guide as a resource.

There will be time in the labs this week to go over points on how to present the results in your long report. Last week our major aim was for you to understand conceptually the pattern of results in the experiment. At the end of this week's labs, you should have a firm understanding on the most appropriate ways of presenting the results you intend on reporting in your long report.

Use this weeks labs effectively as it is your last chance to consult your tutors on long report material.

Chris

Labs | PSY2011 Home Page | Psychology | Monash University