Statistical Skills Gaps of Professors of Education at U.S. Universities and HBCUs

Figures & data

Table 1 Software package features.

		Offline	User-written
Package	Cost	access	packages
SPSS	[dollar]259.99 for 12 month rental^1	No^6	No
R	Free^2	Yes	Yes
SAS	Free university edition^3	Yes	Yes^4
Stata	Stata/IC [dollar]225 perpetual license^5	Yes	Yes

NOTES:

1 On the Hub (2018).

2 CRAN (2018).

3 SAS (n.d.).

4 (Wicklin 2016).

5 (StataCorp LLC 2018).

6 SPSS can be used temporarily offline through a computer license (IBM 2018).

Table 2 Traits of survey respondents U.S. Sample.

Trait	%
At public university	55
Teaches only undergraduates	10
Most recent degree was in U.S.	98
Native English speaker	91
Teacher educator	44
Expected 2+ publications per year	38
No publication expectation	21

Table 3 Software used most often for statistical analyses U.S. sample.

Software	%
Excel	18
SPSS	51
SAS	2
Stata	3
R	1
Other	3
None (Rarely use statistical software)	23

Note. Percentages add to 101% due to rounding.

Table 4 Proportions reporting software skill and need in U.S. sample.

Software	Skill	Need	Gap	${\mathrm{X}}^2$	p
R	0.05	0.24	0.19	21.59	<0.001
Stata	0.12	0.19	0.17	2.85	0.092
Excel	0.95	0.88	0.07	2.93	0.097
SPSS	0.74	0.77	−0.03	0.60	0.438
SAS	0.22	0.26	−0.04	1.47	0.225

NOTE: Proportions indicate moderate or advanced skill or need. All ${\mathrm{X}}^2$. statistics have 1 degree of freedom. Sorted by gap.

Fig. 1 Percent of faculty who have used procedure in last 5 years in general U.S. sample. Percent who do not have help available to them with the procedure.

Table 5 Proportions reporting methodology skill and need in U.S. sample.

Methodology	Skill	Need	Gap	${\mathrm{X}}^2$	p
Structural equation modeling (SEM)	0.28	0.54	0.26	23.71	<0.001
Propensity score matching	0.19	0.42	0.23	21.44	<0.001
Hierarchical linear modeling (HLM)	0.32	0.55	0.23	16.25	<0.001
Item response theory (IRT)	0.38	0.53	0.15	6.93	0.008
Logistic regression	0.46	0.59	0.13	8.58	0.003
Exploratory factor analysis (EFA)	0.47	0.59	0.12	11.22	0.001
Multiple regression	0.61	0.65	0.04	2.02	0.155
ANCOVA	0.59	0.63	0.04	0.78	0.376
Cronbach’s Alpha	0.64	0.65	0.01	0.00	0.999
ANOVA	0.76	0.68	−0.08	2.55	0.110
Correlation	0.75	0.65	−0.10	4.53	0.033
t-test	0.82	0.71	−0.11	6.45	0.011
Summary statistics	0.76	0.61	−0.15	10.24	0.001

NOTE: Proportions indicate moderate or advanced skill or need. All ${\mathrm{X}}^2$ statistics have 1 degree of freedom. Sorted by gap.

Table 6 What prevents respondent from getting more statistics training U.S. sample.

Limiting factor	%
Lack of need or interest	39
Lack of confidence	6
Lack of time for training	53
Lack of opportunity	23
Lack of funds for training	32
Lack of funds for software	23

Table 7 Desired modes for statistics training U.S. sample.

Learning mode	%
Face-to-face workshop	71
Online workshop	33
Face-to-face course	25
Online credit-bearing course	9

Table 8 Scale statistics: U.S. general sample.

				Scale	Scale
	Cronbach’s	Scale	Scale	possible	possible
Scale	Alpha	M	SD	min	max
Software Proficiency	0.77	11.52	2.54	5	15
Need for Software Skills	0.75	12.42	3.1	5	15
Statistical Methods Proficiency	0.95	35.11	8.82	13	39
Use of Statistical Methods	0.91	20.82	4.05	13	39
Need for Statistical Method Skills	0.97	37.65	11.13	13	39
Available Help for Statistical Methods	0.98	28.15	17.77	13	39

Table 9 Traits of survey respondents HBCU sample.

Trait	%
At public university	35
Teaches only undergraduates	18
Most recent degree was in U.S.	96
Native English speaker	91
Teacher educator	41
Expected 2+ publications per year	29
No publication expectation	10

Table 10 Software used most often for statistical analyses HBCU sample.

Software	%
Excel	16
SPSS	54
SAS	0
Stata	3
R	0
Other	2
None (Rarely use statistical software)	24

NOTE. Percentages add to 99% due to rounding.

Table 11 Proportions reporting software skill and need in HBCU sample.

Software	Skill	Need	Gap	${\mathrm{X}}^2$	p
R	0.06	0.24	0.18	14.82	<0.001
SPSS	0.71	0.89	0.18	14.40	<0.001
SAS	0.25	0.41	0.16	9.14	0.003
Stata	0.09	0.22	0.13	10.17	0.001
Excel	0.88	0.94	0.06	6.31	0.012

NOTE. Proportions indicate moderate or advanced skill or need. All ${\mathrm{X}}^2$ statistics have 1 degree of freedom. Sorted by gap.

Fig. 2 Percent of faculty who have used procedure in last 5 years in HBCU sample. Percent who do not have help available to them with the procedure.

Table 12 Proportions reporting methodology skill and need in HBCU sample.

Methodology	Skill	Need	Gap	${\mathrm{X}}^2$	p
Propensity score matching	0.11	0.47	0.36	23.12	<0.001
Item response theory (IRT)	0.28	0.58	0.30	22.05	<0.001
Structural equation modeling (SEM)	0.24	0.54	0.30	18.28	<0.001
Hierarchical linear modeling (HLM)	0.27	0.52	0.25	16.62	<0.001
Cronbach’s alpha	0.46	0.70	0.24	19.44	<0.001
Exploratory factor analysis (EFA)	0.34	0.57	0.23	14.47	<0.001
Logistic regression	0.46	0.65	0.19	7.95	0.005
Multiple regression	0.63	0.79	0.18	10.41	0.001
ANCOVA	0.60	0.71	0.11	5.87	0.015
ANOVA	0.68	0.78	0.10	7.34	0.007
t-test	0.71	0.79	0.08	2.80	0.094
Correlation	0.67	0.72	0.05	4.69	0.030
Summary statistics	0.68	0.58	−0.10	1.63	0.202

Note. Proportions indicate moderate or advanced skill or need. All ${\mathrm{X}}^2$ statistics have 1 degree of freedom. Sorted by gap.

Table 13 What prevents respondent from getting more statistics training HBCU sample.

Limiting factor	%
Lack of need or interest	24
Lack of confidence	15
Lack of time for training	51
Lack of opportunity	31
Lack of funds for training	41
Lack of funds for software	36

Table 14 Desired modes for statistics training HBCU sample.

Learning mode	%
Face-to-face workshop	74
Online workshop	36
Face-to-face course	28
Online credit-bearing course	16

Table 15 Scale statistics: HBCU sample.

				Scale	Scale
	Cronbach’s	Scale	Scale	Possible	Possible
Scale	Alpha	M	SD	Min	Max
Software Proficiency	0.71	11.02	2.38	5	15
Need for Software Skills	0.77	12.53	3.07	5	15
Statistical Methods Proficiency	0.95	32.32	8.54	13	39
Use of Statistical Methods	0.88	21.95	3.46	13	39
Need for Statistical Method Skills	0.97	34.99	11.25	13	39
Available Help for Statistical Methods	0.97	25.59	15.53	13	39

On the Hub. (2018), “IBM SPSS Statistics Faculty Pack 25.” https://estore.onthehub.com/WebStore/OfferingsOfMajorVersionList.aspx?pmv=ef59e1db-5e83-e711-80f3-000d3af41938&cmi_mnuMain=ed6ad73c-7bc7-e011-ae14-f04da23e67f6&utm_source=LandingPage-SPSS-Statistics-faculty&utm_medium=LandingPage-b&utm_campaign=SPSS.

 Google Scholar

CRAN. (2018), “The Comprehensive R Archive Network.” Available at https://cran.r-project.org/.

 Google Scholar

Wicklin, R. (2016), “Writing Packages: A New Way to Distribute and Use SAS/IML Programs.” Available at https://support.sas.com/resources/papers/proceedings16/SAS4201-2016.pdf.

 Google Scholar

StataCorp LLC. (2018), “Stata Campus GradPlan.” Available at http://www.stata.com/order/new/edu/gradplans/campus-gradplan/.

 Google Scholar

IBM. (2018), “What is a Commuter License?” Available at https://www-01.ibm.com/support/docview.wss?uid=swg21482911.

 Google Scholar