Cell phone text messaging patterns and data quality issues

¶ … SPSS

How did you treat missing or oddly coded data, and outliers?

The SPSS function for Missing Data was used to identify any outliers or oddly coded data. The percent of missing data is high at 47.2% or 1063 potential responses to the question. The number of extremes or outliers in the high range is 220. Because the numbers are high, it would be useful to look at the raw data to determine how these answers were coded. This question is an interesting one and the pattern of responses suggests that additional analysis with this variable could reveal relationships with other factors.

How would you summarize the descriptive statistics for each variable?

The standard error (SE) indicates the reliability of the mean. When the standard error is small, it indicates that the sample mean will actually provide a more accurate picture of the actual population mean. The larger the sample size, the smaller the standard error typically is, however, the standard deviation is not affected by sample size in this direct way. The mean for this sample is 61.38, while the standard error of the mean is 5.330. The standard error indicates that the sample mean is not very close to the true mean of the overall population. The standard deviation is relatively large (183.775) which indicates that there is considerable variation in the sample.

Looking at the skewness and kurtosis of the data indicates that it is non-normal. Estimating skewness by establishing a range of twice the standard error (.071 x 2 = 0.142) in a negative and positive direction demonstrates that the data is skewed. Similarly, establishing a range of normality to estimate the kurotsis by multiplying the standard error of Kurtosis by 2 (.142 x 2 = 02.84) and showing minus that value to plus that value.

How would you state a potential null and alternative hypotheses that might be of interest and important to the test, based on your discoveries in this data?

Ho = The number of text messages received on average each day is not related to the number of moving vehicle tickets received by drivers in the study.

Ha = The number of text messages received on average each day is related to the number of moving vehicle tickets received by drivers in the study.

DESCRIPTIVES VARIABLES=q20

/STATISTICS=MEAN SUM STDDEV VARIANCE RANGE MIN MAX SEMEAN KURTOSIS SKEWNESS.

Descriptives

Notes

Output Created

07-AUG-2015 10:17:00

Comments

Input

Data

/Users/Downloads/171243_May_2010_Cell_Phones.sav

Active Dataset

DataSet1

Filter

one>

Weight

one>

Split File

one>

N of Rows in Working Data File

Missing Value Handling

Definition of Missing

User defined missing values are treated as missing.

Cases Used

All non-missing data are used.

Syntax

DESCRIPTIVES VARIABLES=q20

/STATISTICS=MEAN SUM STDDEV VARIANCE RANGE MIN MAX SEMEAN KURTOSIS SKEWNESS.

Resources

Processor Time

00:00:00.02

Elapsed Time

00:00:00.00

[DataSet1] / Users/Downloads/171243_May_2010_Cell_Phones.sav

Descriptive Statistics

N

Range

Minimum

Maximum

Sum

Mean

Std. Deviation

Statistic

Statistic

Statistic

Statistic

Statistic

Statistic

Std. Error

Statistic

Q20. On an average day, about how many text messages do you send and receive on your cell phone?

0

72991

61.39

5.330

Valid N (listwise)

Descriptive Statistics

Variance

Skewness

Kurtosis

Statistic

Statistic

Std. Error

Statistic

Std. Error

Q20. On an average day, about how many text messages do you send and receive on your cell phone?

33773.093

4.280

.071

18.173

.142

Valid N (listwise)

DESCRIPTIVES VARIABLES=q20

/STATISTICS=MEAN SUM STDDEV VARIANCE RANGE MIN MAX SEMEAN KURTOSIS SKEWNESS.

Descriptives

Notes

Output Created

07-AUG-2015 10:17:00

Comments

Input

Data

/Users/Downloads/171243_May_2010_Cell_Phones.sav

Active Dataset

DataSet1

Filter

one>

Weight

one>

Split File

one>

N of Rows in Working Data File

Missing Value Handling

Definition of Missing

User defined missing values are treated as missing.

Cases Used

All non-missing data are used.

Syntax

DESCRIPTIVES VARIABLES=q20

/STATISTICS=MEAN SUM STDDEV VARIANCE RANGE MIN MAX SEMEAN KURTOSIS SKEWNESS.

Resources

Processor Time

00:00:00.02

Elapsed Time

00:00:00.00

[DataSet1] / Users/Downloads/171243_May_2010_Cell_Phones.sav

Descriptive Statistics

N

Range

Minimum

Maximum

Sum

Mean

Std. Deviation

Statistic

Statistic

Statistic

Statistic

Statistic

Statistic

Std. Error

Statistic

Q20. On an average day, about how many text messages do you send and receive on your cell phone?

0

72991

61.39

5.330

Valid N (listwise)

Descriptive Statistics

Variance

Skewness

Kurtosis

Statistic

Statistic

Std. Error

Statistic

Std. Error

Q20. On an average day, about how many text messages do you send and receive on your cell phone?

33773.093

4.280

.071

18.173

.142

Valid N (listwise)

MVA VARIABLES=q20

/LISTWISE.

MVA

Notes

Output Created

07-AUG-2015 10:26:38

Comments

Input

Data

/Users/gigi/Downloads/171243_May_2010_Cell_Phones.sav

Active Dataset

DataSet1

Filter

one>

Weight

one>

Split File

one>

N of Rows in Working Data File

Syntax

MVA VARIABLES=q20

/LISTWISE.

Resources

Processor Time

00:00:00.03

Elapsed Time

00:00:00.00

Univariate Statistics

N

Mean

Std. Deviation

Missing

No. Of Extremesa

Count

Percent

Low

High

q20

61.39

47.2

0

a. Number of cases outside the range (Q1-1.5*IQR, Q3 + 1.5*IQR).

Summary of Estimated Means

q20

Listwise

61.39

All Values

61.39

Summary of Estimated Standard Deviations

q20

Listwise

All Values

Listwise Statistics

Listwise Means

Number of cases q20

61.39

Listwise Covariances

q20

q20

33773.093

Listwise Correlations

q20

q20

1

GET

FILE='/Users/Downloads/May_2010_Cell_Phones.sav'.

DATASET NAME DataSet1 WINDOW=FRONT.

EXAMINE VARIABLES=q20

/PLOT BOXPLOT STEMLEAF

/COMPARE GROUPS

/STATISTICS NONE

/CINTERVAL 95

/MISSING LISTWISE

/NOTOTAL.

Explore

Notes

Output Created

07-AUG-2015 11:15:28

Comments

Input

Data

/Users/Downloads/May_2010_Cell_Phones.sav

Active Dataset

DataSet1

Filter

one>

Weight

one>

Split File

one>

N of Rows in Working Data File

Missing Value Handling

Definition of Missing

User-defined missing values for dependent variables are treated as missing.

Cases Used

Statistics are based on cases with no missing values for any dependent variable or factor used.

Syntax

EXAMINE VARIABLES=q20

/PLOT BOXPLOT STEMLEAF

/COMPARE GROUPS

/STATISTICS NONE

/CINTERVAL 95

/MISSING LISTWISE

/NOTOTAL.

Resources

Processor Time

00:00:01.25

Elapsed Time

00:00:01.00

[DataSet1] / Users/Downloads/May_2010_Cell_Phones.sav

Case Processing Summary

Cases

Valid

Missing

Total

N

Percent

N

Percent

N

Percent

Q20. On an average day, about how many text messages do you send and receive on your cell phone?

52.8%

47.2%

Q20. On an average day, about how many text messages do you send and receive on your cell phone?

Q20. On an average day, about how many text messages do you send and rec

Frequency Stem & Leaf

30.00 0 .

9.00 0 . 888

10.00 1 . 222

30.00 1 . 5555555555&

.00 1 .

.00 1 .

86.00 2 .

.00 2 .

15.00 2 . 55555

.00 2 .

.00 2 .

34.00 3 .

.00 3 .

5.00 3 . 55

.00 3 .

.00 3 .

21.00 4 .

.00 4 .

1.00 4 . & 220.00 Extremes (>=50)

Stem width: 10

Each leaf: 3 case(s) & denotes fractional leaves.

EXAMINE VARIABLES=q20

/PLOT BOXPLOT STEMLEAF NPPLOT

/COMPARE GROUPS

/STATISTICS NONE

/CINTERVAL 95

/MISSING LISTWISE

/NOTOTAL.

Explore

Notes

Output Created

07-AUG-2015 11:21:15

Comments

Input

Data

/Users/Downloads/May_2010_Cell_Phones.sav

Active Dataset

DataSet1

Filter

one>

Weight

one>

Split File

one>

N of Rows in Working Data File

Missing Value Handling

Definition of Missing

User-defined missing values for dependent variables are treated as missing.

Cases Used

Statistics are based on cases with no missing values for any dependent variable or factor used.

Syntax

EXAMINE VARIABLES=q20

/PLOT BOXPLOT STEMLEAF NPPLOT

/COMPARE GROUPS

/STATISTICS NONE

/CINTERVAL 95

/MISSING LISTWISE

/NOTOTAL.

Resources

Processor Time

00:00:00.91

Elapsed Time

00:00:01.00

Case Processing Summary

Cases

Valid

Missing

Total

N

Percent

N

Percent

N

Percent

Q20. On an average day, about how many text messages do you send and receive on your cell phone?

52.8%

47.2%

Tests of Normality

Kolmogorov-Smirnova

Shapiro-Wilk

Statistic

df

Sig.

Statistic

df

Sig.

Q20. On an average day, about how many text messages do you send and receive on your cell phone?

.378

.000

.349

.000

a. Lilliefors Significance Correction

Q20. On an average day, about how many text messages do you send and receive on your cell phone?

Q20. On an average day, about how many text messages do you send and rec Stem-and-Leaf Plot

Frequency Stem & Leaf

30.00 0 .

9.00 0 . 888

10.00 1 . 222

30.00 1 . 5555555555&

.00 1 .

.00 1 .

86.00 2 .

.00 2 .

15.00 2 . 55555

.00 2 .

.00 2 .

34.00 3 .

.00 3 .

5.00 3 . 55

.00 3 .

.00 3 .

21.00 4 .

.00 4 .

1.00 4 . & 220.00 Extremes (>=50)

Stem width: 10

Each leaf: 3 case(s) & denotes fractional leaves.

• Explore one of the questions with a non-scale measurement in the survey and report its level of measurement (either nominal or ordinal) and measure of central tendency and variability. If there are missing data, outliers, or oddly coded data (Don't Knows or Refusals, et cetera), address them to report how you ensured these issues did not affect your estimates of descriptive statistics. This analysis should be done using the SPSS Frequencies command under Analyze.

Reference

de Mast, J., & Kemper, B.P.H. (2009). Principles of exploratory data analysis in problem solving: What can we learn from a well-known case? Quality Engineering, 21(4), 366 -- 375.