You are on page 1of 8

RESEARCH QUESTIONS

 Is the mean IQ equal for Indian or Native American, Asian or Pacific Islander, White

and the rest?

 Is the mean annual income of the participant’s household equal for Indian or Native

American, Asian or Pacific Islander, White and the rest?

 Is the mean sense of school belonging equal for Indian or Native American, Asian or

Pacific Islander, White and the rest?

MODIFIED METHODS

PARTICIPANTS

The participants in this study were Indian or Native American children with a mean age of

13.63 and a standard deviation of 0.55, Asian or Pacific Islander children with a mean age of

13.56 and a standard deviation of 0.52, White children with a mean age of 13.58 and a

standard deviation of 0.52, and children from the rest of the races with a mean age of 13.56

and a standard deviation of 0.55. In all the 4 groups, the children were aged between 12 and

14 and they were taken from a population with a mean age of 13.57 and a standard deviation

of 0.53.

DATA ANALYSIS PLAN

The statistical analysis software that was used to conduct the analysis was SPSS.

One – way ANOVA

Before starting to conduct the tests, the distribution of the data was looked at. Thus,

histograms were plotted, and the mean and standard deviation of data values were calculated.

After looking at the distributions, a homogeneity test was conducted, and this determined the

kind of post hoc test to conduct. Thus, if the test of homogeneity was to return equal
variances, the post hoc test to be conducted was going to be Tukey test. If the test of

homogeneity showed that the variances were unequal, then the post hoc test to be conducted

was going to be Dunnett T3 test.

RESULTS

Normality Test

The following histograms and normal Q – Q plots show the results of normality test

Figure 1: Histogram showing frequency of IQ.


From Figure 1, the histogram is bell shaped hence it showed that IQ was normally distributed
hence no further normality tests were required.
Figure 2: Histogram showing participants household annual income frequency.

From figure 2 above, the histogram was skewed to the right hence did not show normal
distribution. But normality was assumed since the sample size was large thus, next it was
checked that each of the 4 categories had a sample size that was greater than or equal to 30.
The descriptive statistics showed this was the case hence normality was assumed. This was
also the case with sense of belonging as shown in figure 3 below. Sense of belonging was
skewed to the left. But, since all the categories checked have a large sample size, normality
was still assumed.

Figure 3: Showing frequency of sense of belonging.


The descriptive statistics were summarized in table 1 below.
Table 1: Showing descriptive statistics of the 4 independent variables.

IQ Household School belonging


Independent Variable N M SD M SD M SD
Indian or Native American 32 95.16 12.04 30.63 19.20 2.37 1.07
Asian or Pacific Islander 100 102.85 17.92 57.36 71.40 2.91 0.69
White 881 105.48 12.91 51.57 42.56 2.90 0.72
The rest 541 94.96 14.48 36.08 32.97 2.77 0.76
Total 1556 101.43 14.68 46.11 42.71 2.84 0.74

One – way ANOVA


The results of the one-way ANOVA were summarized in the following table

Table 2: Summarized one – way ANOVA results.

Independen
2
t Variable df F η p
[3,1552
IQ ] 67.339 0.115 <.001
[3,1552
Household ] 19.15 0.036 <.001
SCHLBELON [3,1552
G ] 8.611 0.016 <.001

Pairwise Comparisons (Post Hoc Dunnett T3)

After conducting the ANOVA analysis, a pairwise comparison was conducted, and the results

were summarized in the table below.

Table 3: Showing pairwise comparison results.

IQ Household School belonging


Group 1 Group 2 p p p
Indian or Native American Asian or Pacific Islander .042 .006 0.058
<.00
Indian or Native American White 1 <.001 0.047
Indian or Native American The rest 1 .600 0.232
Asian or Pacific Islander White .640 .964 1
<.00
Asian or Pacific Islander The rest 1 .025 0.317
<.00
White The rest 1 <.001 0.004
DISCUSSION

One – way ANOVA

There was a significant effect of Race on IQ at the 5% significance level, F (3,1552) =

67.339, p < .001, η2 =.115. The effect of race on IQ was moderately high based on the effect

size. Therefore, at least one mean IQ was different based on race.

There was a significant effect of Race on annual household income at the 5% significance

level, F (3,1552) = 19.15, p < .001, η2 =.036. The effect of Race on annual household income

was small based on the effect size. Therefore, at least one mean annual household income

was different based on race.

There was a significant effect of Race on sense of belonging at the 5% significance level, F

(3,1552) = 8.611, p < .001, η2 =.016. The effect of Race on sense of belonging was small

based on the effect size. Therefore, at least one mean sense of belonging was different based

on race.

Thus, to find out which means were different, a post hoc test was conducted. Before

conducting a post hoc test, a test of homogeneity of variances was conducted where it was

found that the variances were unequal in all the cases studied. Thus Dunnett’s T3 pairwise

comparison was used.

Post Hoc Test

For IQ, the mean IQ for Indian Native American was found to be different from the mean IQ

for Asian or Pacific Islander, p = .042, and White children, p < .001. Then, the mean IQ for

the rest of the children was found to be different from the mean IQ for Asian or Pacific

Islander, p < .001, and White children, p < .001.


For annual household income, the mean annual household income for Indian Native

American was found to be different from the mean annual household income for Asian or

Pacific Islander, p =.006, and White children, p < .001. Then, the mean annual household

income for the rest of the children was found to be different from the mean annual household

income for Asian or Pacific Islander, p = .025, and White children, p < .001.

For sense of belonging, the mean sense of belonging for Indian Native American was found

to be different from the mean sense of belonging for White children, p = .047. Then, the mean

sense of belonging for the rest of the children was found to be different from the mean sense

of belonging for White children, p =.004.

The post hoc test was conducted at 5% significance level.


SPSS SYNTAX

 First a new variable was created in excel (ANAWHITEAPI ) where an IF function

was used to set the values for ANA = 1 to 1, for API = 1 to 2 and for WHITE = 1 to 3

and the rest to 4.

 The data was then imported in SPSS, (File → Import Data →CSV Data)

 Then descriptive statistics for age: ( Analyze→ Descriptive Statistics→ Explore) In

explore, dependent list was age and factor was the new variable created

ANAWHITEAPI. In statistics, descriptive statistics was check marked and then Ok

was pressed.

 Then, Then descriptive statistics for IQ, HOUESHOLD and SCHLBELONG:

( Analyze→ Descriptive Statistics→ Explore) In explore, dependent list was IQ,

HOUSEHOLD and SCHLBELONG and cases were labelled by ANAWHITEAPI. In

statistics, descriptive statistics was check marked and in plots histogram was

checked then Ok was pressed.

 After, ANOVA analysis was conducted.

 Thus, Analyze→ Compare Means → One−Way Anova

 Dependent list ¿ IQ, HOUSEHOLD and SCHLBELONG

 Factor ¿ ANAWHITEAPI

 Post hoc, Tukey and Dunnett’s T3 were checked.

 Options descriptives and homogeneity of variance test were checked.

 Estimate effect size was checked.

 Then, OK was pressed.

You might also like